ArticlePDF AvailableLiterature Review

The Molecular basis of Lactose Intolerance


Abstract and Figures

A staggering 4000 million people cannot digest lactose, the sugar in milk, properly. All mammals, apart from white Northern Europeans and few tribes in Africa and Asia, lose most of their lactase, the enzyme that cleaves lactose into galactose and glucose, after weaning. Lactose intolerance causes gut and a range of systemic symptoms, though the threshold to lactose varies considerably between ethnic groups and individuals within a group. The molecular basis of inherited hypolactasia has yet to be identified, though two polymorphisms in the introns of a helicase upstream from the lactase gene correlate closely with hypolactasia, and thus lactose intolerance. The symptoms of lactose intolerance are caused by gases and toxins produced by anaerobic bacteria in the large intestine. Bacterial toxins may play a key role in several other diseases, such as diabetes, rheumatoid arthritis, multiple sclerosis and some cancers. The problem of lactose intolerance has been exacerbated because of the addition of products containing lactose to various foods and drinks without being on the label. Lactose intolerance fits exactly the illness that Charles Darwin suffered from for over 40 years, and yet was never diagnosed. Darwin missed something else--the key to our own evolution--the Rubicon some 300 million years ago that produced lactose and lactase in sufficient amounts to be susceptible to natural selection.
Content may be subject to copyright.
Black plate (2,1)
Science Progress
One hundred years of reporting science
Contents Volume 88 | Part 3 | 2005
Extracellular sensors and extracellular alarmones, which
permit cross-talk between organisms, determine the levels
of alkali tolerance and trigger alkali-induced acid
sensitivity in
Escherichia coli
The molecular basis of lactose intolerance 157
Front cover: The idea that bacteria in the gut can release toxins is over 100 years old.
Elie Metchnikoff was a founder of modern immunology, discovering phagocytes for
which he was awarded one of the earliest Nobel Prizes with Ehrlich in 1902. A
staggering 4000 million people cannot digest lactose, the sugar in milk, properly. The
symptoms of lactose intolerance are caused by gases and toxins produced by anaerobic
bacteria in the large intestine. The molecular basis for this intolerance is discussed in
pages 157–202.
Black plate (2,1)
Michael Abraham
(Physical Sciences)
Department of Chemistry
University College London
20 Gordon Street
London WC1H 0AJ
Robin Rowbury
(Life Sciences)
Department of Biology
University College London
Darwin Building, Gower Street
London WC1E 6BT, UK
Telephone +44(0)20 7387 7050
Fax +44(0)20 7380 7096
Sara Nash
Science Reviews
PO Box 314
St. Albans
Herts AL1 4ZG, UK
Telephone/Fax +44(0)1727 764601
P.H. Clarke FRS
D. Lewis FRS
D. Phillips OBE
P.J.B. Slater
J.N. Ziman FRS
Klaus Suhling
Department of Physics
King’s College London
Strand WC2R 2LS, UK
Telephone +44(0)20 7848 2119
Mark Green
Department of Physics
King’s College London
Strand WC2R 2LS, UK
Telephone +44(0)20 7848 2121
Peter W. Atkins, Lincoln College, Oxford, UK
R.J. (Sam) Berry, University College, London, UK
Patrick Dowling, University of Surrey, UK
J. Michael Bishop, UC Medical Center, San Francisco, USA
Peter Edwards, FRS, School of Chemistry, University of Birmingham, UK
Baroness Susan Greenfield, Royal Institution, London, UK
Lauri D. Hall, School of Clinical Medicine, University of Cambridge, UK
Dame Julia Higgins, FRS, Imperial College London, UK
ean-Marie Lehn, Nobel Laureate, Institut le Bel, Universite
´Louis Pasteur, France
The Lord Lewis, FRS, Robinson College, Cambridge, UK
erry Mansfield, FRS, School of Biological Science, University of Lancaster, UK
Hiroshi Masuhara, Faculty of Engineering, Osaka University, Japan
Brian Spratt, FRS, School of Medicine, Imperial College London, UK
George Stewart, University of Brisbane, Australia
Black plate (133,1)
Extracellular sensors and
extracellular alarmones, which
permit cross-talk between
organisms, determine the levels of
alkali tolerance and trigger alkali-
induced acid sensitivity in
Escherichia coli
For several stress responses in Escherichia coli, switching on involves
conversion by the stress of an extracellular stress sensor (an extracellular
sensing component, ESC) to an extracellular induction component (EIC),
the latter functioning as an alarmone and inducing the response. The aim of
this study was to establish whether alkali tolerance induction at pH 9.0,
alkali sensitisation induced at pH 5.5 and the acid sensitisation induced at
pH 9.0 involve sensing of pH changes by ESCs. The techniques involved
made use of studies with cell-free culture filtrates. With respect to the
inducible responses under test, these filtrates were prepared either from
induced or uninduced cultures and filtrates from uninduced cultures were also
activated in vitro, by the pH stress, in the absence of bacteria. Tests were
then made to examine whether EICs (known to be needed for all these
systems) are formed by activation, at the appropriate pH values, of filtrates
from pH 7.0-grown cultures (i.e. uninduced culture filtrates); appearance of
an EIC on activation would indicate the presence in the uninduced culture
filtrate of an ESC. The studies showed that all three systems use ESCs to
detect pH changes. Tests involving attempted enzymic and physical inactiva-
tion of the ESCs, and attempted removal of the ESCs by dialysis, showed
that the ESC involved in alkali sensitisation is a small very heat-resistant
protein. Strikingly, protease only partially inactivated the ESCs needed for
alkali tolerance induction and for acid sensitisation; each system may be
Science Progress (2005), 88(3), 133–156 133
Black plate (134,1)
complex, involving both protein and non-protein (RNA?) ESCs, although
other explanations are possible. It was also established that appropriate
killed cultures can induce all three responses when incubated with pH 7.0-
grown living cultures. The occurrence of ESCyEIC pairs for these three
responses has led to the evolution of early warning systems for each, the
diffusibility of the EICs, and their interaction with non-producers, allowing
them to act pheromonally, inducing sensitive organisms to stress tolerance,
prior to exposure to stressor.
Keywords:acid sensitivity induction; alkali sensitisation; alkali tolerance
induction; ASI; cross-talk; Escherichia coli; extracellular alarmones;
extracellular sensing components; extracellular induction components;
intercellular communication; modification of sensors; novel sensing
Robin Rowbury is Emeritus Professor of Microbiology at University College
London, and for further information on this work can be reached at Margaret Goodson worked for
many years, prior to retirement, at UCL with Prof Rowbury’s group, and
the researches of the group were, for more than 20 years on factors influencing
bacterial survival, particularly in the aquatic environment. In recent years,
work was on how enterobacteria could enhance survival, in response to
chemical, physical and biological stressors, by switching-on inducible stress
responses. These studies led to the finding of how such responses were
triggered by extracellular induction components (EICs), and then how the
latter were produced by the activation, by appropriate stressors, of extra-
cellular sensors (extracellular sensing components; ESCs). It has now been
established that such pairs of extracellular components (ECs) allow for early
warning against potentially lethal stressors, organisms being able, in response
to diffusible EIC alarmones, to induce tolerance responses prior to stressor
1. Introduction
1.1 Stresses affecting polluting, contaminating and infecting
It is now well established that the ability of bacteria to tolerate
stressing agents and conditions is a major factor in their ability to
survive in numerous natural and man-made situations and to grow
when the stressing condition has passed
. In the natural aquatic
environment, polluting enterobacteria can be subjected to many
chemical, physical and biological stresses
and the levels of their
134 Robin J. Rowbury and Margaret Goodson
Black plate (135,1)
inherent and induced stress tolerances determine whether or not
they survive
2,7 – 11
. Exposure to alkaline pH values in natural waters
is common. First, alkalinity can derive from the entry into water
courses of wastes from chemical plants or poultry processing plants
or of basic sewage or slurries from the outfall at some agricultural
sites, such slurries often containing ammonia or other alkaline
. Additional alkalinisation of waters can occur following
run-offs from some basic soils. Contaminating enterobacteria can
also face alkalinity in some foods, with the high pH of egg white
being of particular significance
and exposure to alkalinity can
occur at some stages of food processing
. Also, infecting organ-
isms can be subjected to alkalinity in some parts of the intestine and
in the phagolysosome, where a high pH phase follows the initial
. Exposures to acidity in the natural environment
due for example to the precipitation of acid rains or acid snow, or
following the entry into natural waters of acidic chemical wastes,
are common, whilst low pH exposures often occur in foods and in
food processing. In the body, infecting organisms frequently face
acidity e.g. in the stomach and phagolysosome
. The studies
described involve mainly responses to acidity and alkalinity, but
bacteria can be exposed to numerous other chemical stresses
including metal ions, detergents, oxidative components, high salt
concentrations and other osmotic stressors, as well as to stressing
physical conditions such as heat, cold and irradiation stresses. The
ability of organisms to survive the above stresses is initially
dependent on inherent tolerances processes
, but if these are not
to be overwhelmed when stress levels increase, induction of new
responses must occur and many inducible stress tolerance responses
have been observed
. Although exposure to a stress generally
induces tolerance to the same stress, it can also induce cross-
tolerance responses to other stresses
. Because inducible toler-
ance can be critical for survival of organisms in numerous loca-
tions, the mechanisms by which induction occurs are of interest.
The above refers to induction of stress tolerance. It has also been
established that inducible responses to certain stresses can also lead
to sensitivity to other stresses
and were the conditions which
induced such sensitivity to be followed by challenge with the
appropriate second stress, this could be lethal. As stated above,
amongst the stresses faced by organisms are extremes of alkalinity
and only if organisms induce alkali tolerance
can they survive the
highest pH values. As with other stresses, alkalinity, when occur-
ring in many locations, builds up gradually, from levels which are
resistable by inherent tolerance processes only, to those which are Extracellular sensors and extracellular alarmones 135
Black plate (136,1)
lethal to nearly all organisms, unless they have induced alkali
tolerance. Such tolerance is induced at around pH 8.5 – 9.0, but
at these pH values, organisms also become sensitised to acid
process which involves regulatory components quite distinct from
those needed for alkali tolerance induction, and is termed ASI (acid
sensitivity induction). Alkali tolerance is switched-on at mildly
alkaline pH values, but strikingly, organisms become sensitised to
alkali if exposed to mild acidity (especially pH 5.5 – 6.0).
1.2 Mechanisms of initiation of stress tolerances
Many inducible processes in enterobacteria are believed to be
switched-on by the inducer (stimulus) crossing the outer membrane
and interacting in the periplasm with a sensing group
, often part
of an integral cytoplasmic membrane protein
, interaction leading
to the production of a signal which directly or indirectly switches-
on the response. Other systems involve the stimulus crossing both
membranes before interacting with a cytoplasmic sensor
. The
present study re-inforces other work
which has established that
many stress responses need extracellular components which switch
them on; these extracellular induction components (EICs) are not
only required for response induction in the stressed cells, but can
also diffuse to regions not reached by the stress and induce
tolerance in unstressed organisms i.e. the production of EICs in
the region of the stress allows cross-talk between organisms, the
intercellular communication which results then allowing unstressed
organisms to show response induction prior to stress exposure.
Thus the EICs are acting as both pheromones and alarmones. For
some of the above responses, stresses are sensed by extracellular
sensing components (ESCs;
25 – 28
) which are converted by the
stresses into the above extracellular induction component (EIC)
alarmones which interact with the organisms to switch-on
. The findings here show that some ESCs can detect
rising pH and that their activation to EICs leads to the switching-
on of both alkali tolerance and acid sensitivity; an ESC is also
shown to detect the acidification which leads to induction of alkali
1.3 Killed cultures can confer stress responses on living
Strikingly, stress-responding ECs show highly distinctive behaviour
with respect to their responses to mild stress and to potentially
136 Robin J. Rowbury and Margaret Goodson
Black plate (137,1)
lethal stress. ESCs are exquisitely sensitive to activation (converting
them to EICs) by tiny concentrations of stressors, and EICs can be
converted back to ESCs by low levels of certain stresses. In
contrast, ESCs and EICs are very resistant to irreversible inactiva-
tion by high levels of stressors. Because ESCs and EICs are often
resistant to a range of inhibitory agents and conditions, treatments
which kill organisms do not necessarily inactivate ESCs and EICs.
Such killed cultures might be expected, therefore, to induce stress
responses in living cultures and acid tolerance induction by several
killed cultures has been demonstrated
. Here, appropriate killed
cultures have also been shown to induce, in living cultures, all three
responses under test.
2. Experimental studies of responses to alkali
and acid
2.1 Bacterial strains and culture conditions
The strains used here were strain 1829 (trp,lac)
and strain 1157
(thr,leu,proA2 D[proA,phoE,gpt], his,thi,argE,lacY,galK,xyl,
, both derivatives of Escherichia coli K12. Strain 1829 has
been used previously for alkali tolerance induction
and alkali
studies and is used for that purpose here. Because
strain 1157 shows a particularly marked acid sensitisation (acid
sensitivity induction, ASI) at alkaline pH, it is used for the
experiments on ASI described here. Strains were grown in Oxoid
No. 2 broth (25 g l
) of the stated pH at 37C, initially overnight
in such broth and then diluted 100-fold into fresh broth with
growth to midlog phase. For viable count assessment, samples
were plated on the above broth medium solidified with Difco
granulated agar (20 g l
2.2 Preparation and use of cell-free medium filtrates
Midlog phase cultures grown as described above were filtered
through Acrodisc 0.2 mm pore size filters (Gelman Sciences, Ann
Arbor, MI, USA); plating of filtrates showed that they were
virtually free of viable organisms. The main experiments under-
taken were to demonstrate involvement of ESCs in switching on
of the systems being studied. For alkali tolerance induction,
tolerance is switched on in cultures when pH 7.0-grown organisms
are shifted to pH 9.0
. By analogy with other systems
25 – 28
, it was
likely that there is an ESC present in filtrates from cultures grown Extracellular sensors and extracellular alarmones 137
Black plate (138,1)
at pH 7.0 (pH 7.0 filtrates or uninduced filtrates), which would be
activated at pH 9.0, in the absence of organisms, to the alkali
tolerance inducing EIC. Accordingly, filtrates from pH 7.0-grown
strain 1829 were incubated at pH 9.0 for 30 min and after
neutralisation, activated filtrates were tested for presence of
EIC. Such tests involved incubating pH 7.0-grown strain 1829
culture (1 part) with activated filtrate (1 part) for 60 min at pH
7.0. Organisms were then assessed for increased alkali tolerance
by challenging at pH 11.0 for 5 or 6 min. Where required, the pH
7.0 filtrates were subjected, prior to activation, to a range of
treatments, to ascertain the nature and properties of the ESC.
Tests were with heat (15 min at 75C), with dialysis (against 20 vol
pH 7.0 broth for 18 h at 4C), with protease (1 unit ml
, Sigma
P4531 protease) for 30 min at 30C prior to removal of enzyme by
filtration through glass fibre filters (Gelman Sciences, Ann Arbor,
MI, USA), with RNase (1 unit ml
Sigma R6513 RNase) for
15 min at 30C, followed by removal of enzyme by filtration, as
above or with DNase (1 unit ml
Sigma D4527 DNase) for
15 min at 30C. DNase on beads was not available and, therefore,
the enzyme was not removed after such treatment. Alkali sensi-
tisation occurs when organisms growing at pH 7.0 are shifted to
pH 5.5
. Accordingly, to examine involvement of an ESC in
induction, cell-free filtrates from pH 7.0-grown strain 1829 were
incubated at pH 5.5 for 30 min and, after neutralisation, EIC
assessed by incubating pH 7.0-grown strain 1829 with activated
filtrate (1y1 filtrate to culture) for 60 min at pH 7.0. Incubated
organisms were then examined for sensitisation to alkali by
challenge at pH 10.5 for 8 min. To establish the nature and
properties of the ESC, some of the above filtrates were treated
with heat, dialysis, protease, RNase or DNase, prior to activation,
as described above. Acid sensitivity induction (ASI) is switched on
when cultures of strain 1157 are shifted from pH 7.0 to pH 9.0
Accordingly, to ascertain whether an ESC was involved in
induction, cell-free filtrates from pH 7.0-grown cultures of this
strain were shifted to pH 9.0 for 30 min, in the absence of
organisms. After neutralisation of filtrates, they were incubated
with pH 7.0-grown strain 1157 (1y1 filtrate to culture) for 60 min
at pH 7.0 and organisms were then challenged at pH 3.0 for
7 min. Studies on the nature and properties of the ESC were
undertaken by examining the effects of heat, dialysis, protease,
RNase and DNase on the filtrates as described above. For all the
above studies on ESC- and EIC-containing filtrates, tests were
also made using filtrates from uninduced cultures as controls.
138 Robin J. Rowbury and Margaret Goodson
Black plate (139,1)
2.3 Induction of stress responses by killed cultures
Here, cultures induced for the various responses (cultures being
transferred to inducing conditions either before or after killing) were
killed and then incubated with pH 7.0-grown cultures of strain 1829
(for studies of alkali tolerance induction or alkali sensitisation) or
strain 1157 (for studies of ASI); incubations were for 60 min at pH
7.0 with killed culture (neutralised if required prior to incubation),
mixtures containing one part of living culture and one part of killed
culture. After incubation, the mixtures were challenged under the
appropriate conditions to ascertain whether the tolerance or sensi-
tisation properties had been transferred to the living culture. The
killing conditions were as follows. For strain 1829 to be used for
alkali tolerance induction, culture grown at pH 7.0 was killed by
exposure to UV irradiation for 5 min, with culture at 15 cm from a
Philips TUV 6W tube and then, after killing, activated by exposure
to pH 9.0 for 30 min prior to incubation of neutralised killed culture
with pH 7.0-grown strain 1829 at pH 7.0. For strain 1829 to be used
for alkali sensitisation, killing was by exposure to 70C for 15 min.
Both pH 7.0-grown and pH 5.5-grown cultures were killed, the pH
7.0-grown one being activated at pH 5.5 after killing and both killed
cultures being neutralised prior to incubation with pH 7.0-grown
strain 1829. Finally, pH 7.0-grown strain 1157 used for ASI
experiments was killed by exposure to 65C for 15 min and activated
at pH 9.0 for 30 min. After neutralisation, killed culture was
incubated, as above, with pH 7.0-grown strain 1157 for 60 min at
pH 7.0. All the tested killed cultures contained less than 0.01%
survivors after the killing process.
3. Experimental results
It had already been established that the three inducible processes
described here involve the functioning of EICs for the switching-on
. By analogy with the acid tolerance induction systems and
the alkylhydroperoxide tolerance induction system
25 – 27
, it was also
expected that the three systems studied here would use ESCs as the
components sensing the inducing pH. This has been examined and
established in the sections below.
3.1 An ESC is required for alkali tolerance induction
The alkali tolerance induction response, which is rapidly induced
on transfer of E. coli to pH 9.0, is absolutely dependent on the Extracellular sensors and extracellular alarmones 139
Black plate (140,1)
presence of an EIC in the alkalinised culture. This component,
added in neutralised cell-free filtrates prepared from the alkalinised
culture, can also induce alkali tolerance in E. coli incubated at pH
. In contrast to the above, filtrates from pH 7.0-grown cultures
are ineffective in inducing alkali tolerance in organisms at pH 7.0.
These pH 7.0 filtrates, which are free of viable organisms, do,
however, contain an extracellular component (EC) needed for
alkali tolerance induction, since they can be activated to produce
EIC-containing filtrates. Thus, if cell-free filtrates from pH 7.0-
grown cultures are incubated, without organisms, for 30 min at pH
9.0 and then neutralised, these filtrates gain the ability to induce
alkali tolerance at pH 7.0 (Table 1). Accordingly, the EIC is not
synthesised de novo in cultures incubated at pH 9.0; the pH 7.0
filtrates contain an EC which senses alkaline pH and is converted
by it to an EIC. This agent behaves as a pH sensor i.e.itisan
extracellular sensing component, ESC. As stated, the pH 7.0
cultures contain this ESC, but it is not present in pure broth
(Table 1) i.e. the ESC is probably synthesised de novo by the
organisms at pH 7.0, although these organisms could possibly
form it from a broth component. The alkali tolerance ESC (or
ESCs) present in the pH 7.0 cultures, proved to be rather heat-
resistant, to be somewhat sensitive to RNase and DNase and
partially inactivated by protease (Table 2). On subjecting pH 7.0
filtrates to dialysis against pure pH 7.0 broth, about 30% of the
ESC was removed.
3.2 Modification of the alkali tolerance inducing sensor
during growth at acidic pH
It has been established that the acid tolerance ESC formed at pH
8.0 – 9.0 is markedly different in its activation pH from that arising
at pH 7.0
and the proposal is that such modifications allow the
most efficient functioning of the sensor. The results here show that
the alkali tolerance sensor shows altered activation pH values if
formed at acidic pH. For these studies, organisms of strain 1829
were grown to log-phase at either pH 5.5 or pH 7.0 and ESC-
containing cell-free filtrates prepared. Both filtrates were able to
sense increased pH and showed formation of an EIC resulting from
activation of the ESCs by these increased pH values. Strikingly,
however, the ESC formed at pH 5.5 was markedly activated to EIC
at pH 7.0 7.5 (with slight activation at even pH 6.5), whereas only
slight activation of the pH 7.0 ESC occurred at pH 7.5 (Table 3).
140 Robin J. Rowbury and Margaret Goodson
Black plate (141,1) Extracellular sensors and extracellular alarmones 141
Table 1 Activation of an alkali-detecting sensor involved in alkali tolerance-induction
Filtrate added
to culture
pH for
Culture incubated
with or without
% survival ( SEM),
after alkali challenge,
for organisms incubated
with or without
No filtrate N.A. pH 7.0-grown 0.9 0.12
No filtrate N.A. pH 9.0-grown 20.4 0.68
From pH 7.0 pure broth pH 9.0 pH 7.0-grown 0.4 0.15
From pH 7.0-grown culture pH 7.0 pH 7.0-grown 0.6 0.2
From pH 7.0-grown culture pH 9.0 pH 7.0-grown 15.5 1.8
From pH 9.0-grown culture N.A. pH 7.0-grown 17.60.52
Filtrates were prepared from midlog-phase 1829 culture (or from pure broth) followed by neutralisation, and filtrates were then activated at
the stated pH. Neutralised filtrates were then incubated with midlog-phase 1829 culture (1 part filtrate to 1 part culture) for 60 min at 37C,
washed with pH 7.0 broth and then challenged with alkali (pH 11 for 5 min). N.A. ¼not applicable. SEM ¼standard error of the mean.
Black plate (142,1)
142 Robin J. Rowbury and Margaret Goodson
Table 2 Properties of the alkali-tolerance sensor
Potentially inhibitory
treatments on
pH for
of filtrate
Alkali tolerance of
pH 7.0-grown organisms,
pre-incubated with filtrate;
% survival SEM
after challenge with alkali
None 7.0 0.4 0.1
None 9.0 19.2 3.0
Protease-treatment 9.0 12.0 3.0
75C-treatment 9.0 12.1 4.8
100C-treatment 9.0 7.2 0.6
Dialysis 9.0 12.8 0.25
DNase-treatment 9.0 7.3 0.35
RNase-treatment 9.0 12.5 0.58
Filtrates were prepared from strain 1829 grown to log-phase at 37CinpH7.0
broth and treated (as described in section 2) with P4531 protease on beads, or
with R6513 RNase on beads, or D4527 DNase, or at 75C for 15 min, or at
100C(i.e. in a boiling water bath) for 15 min or by dialysis for 18 h at 4C.
Protease and RNase were removed by passage through Gelman glass fibre
filters and then filtrates were activated at the stated pH followed by neutralisa-
tion where necessary. After incubation of filtrate with 1829 culture (see Table 1
for details), and washing with pH 7.0 broth, challenge of organisms was at pH
11.0 for 5 min.
Table 3 Modification of the alkali-tolerance sensor at pH 5.5
pH for
pH for
% survival ( SEM)
after alkali challenge for
pH 7.0-grown organisms
pre-incubated with filtrate
7.0 7.0 0.83 0.16
7.0 7.5 3.4 0.35
7.0 8.0 11.3 0.71
7.0 9.0 16.9 0.86
5.5 6.5 3.8 1.4
5.5 7.0 8.1 0.97
5.5 7.5 15.8 1.75
5.5 8.0 21.9 3.15
Filtrates were prepared from log-phase 37C cultures of strain 1829 grown at
pH 7.0 and 5.5 followed by filtrate neutralisation. Filtrates were then
incubated for 30 min at 37C at the stated pH values for activation of ESC.
After neutralisation, activated filtrates were incubated with 1829 culture (see
Table 1 legend) followed by washing with pH 7.0 broth and alkali challenge
(pH 11 for 6 min).
Black plate (143,1)
3.3 Alkali sensitisation induced at pH 5.5 is switched-on by
activation of an ESC
Earlier studies showed that the alkali sensitisation induced at pH
5.5 is dependent on an EIC present in the pH 5.5 culture
interacting with the organisms
. This EIC, which proved to
be a very heat-stable low molecular weight protein, can also
induce alkali sensitivity in organisms at pH 7.0. Although pH
5.5-grown culture filtrates contain the above EIC and can, after
neutralisation, induce alkali sensitivity in pH 7.0 cultures, pH 7.0
filtrates are ineffective. These latter filtrates, which are free of
viable cells, do, however, contain an EC involved in sensitisation,
since they can be activated at pH 5.5 in the absence of organisms,
to give filtrates which contain a sensitisation-inducing EIC.
Establishing this is the finding that pH 7.0 filtrates, after exposure
to pH 5.5 for 30 min, can induce sensitisation in pH 7.0 cultures
(Table 4). Pure broth cannot be activated in this way and,
accordingly, the ESC is synthesised by organisms at pH 7.0 or
is formed by them at this pH from a broth component. Tests on
the ESC for alkali sensitisation (Table 4) indicate that this
component is a very heat-stable low molecular weight (dialysable)
protein. The ESC for alkali sensitisation, like that needed for
alkali tolerance induction, occurs in different forms, depending on
the culture conditions. Thus, the ESC formed at pH 8.5 is
activated to EIC at a higher pH value than the ESC formed at
pH 7.0.
3.4 Killed cultures can induce alkali tolerance and alkali
It has previously been found
that killed cultures can induce acid
tolerance in living organisms at pH 7.0, and it appeared that
ESCs or EICs in the killed cultures produced these effects. The
results in Table 5 reveal that ESC-containing cultures killed by
UV-irradiation and then activated to EICs at pH 9.0 can, after
neutralisation, induced alkali tolerance in living cultures at pH
7.0. Additionally, EIC-containing cultures of strain 1829, after
neutralisation and killing at 70C, induced alkali sensitivity in
cultures at pH 7.0 (Table 5) and ESC-containing cultures, killed
at 70C, produced the same effect if activated at pH 5.5 (then
neutralised) following killing (Table 5). Extracellular sensors and extracellular alarmones 143
Black plate (144,1)
144 Robin J. Rowbury and Margaret Goodson
Table 4 An extracellular sensing protein (an ESC) is needed for induction of alkali sensitivity at pH 5.5
used to induce
alkali sensitivity
Filtrate treatment
prior to activation
of filtrate
incubated with
or without filtrate
% survival ( SEM)
for organisms after
challenge at pH
10.5 for 8 min
None N.A. N.A. pH 7.0-grown 23.3 1.28
None N.A. N.A. pH 5.5-grown 5.2 0.59
ESC None At pH 7.0 pH 7.0-grown 26.6 1.5
ESC None At pH 5.5 pH 7.0-grown 4.3 1.3
ESC With protease At pH 5.5 pH 7.0-grown 24.9 4.2
ESC With RNase At pH 5.5 pH 7.0-grown 4.8 0.46
ESC With DNase At pH 5.5 pH 7.0-grown 9.3 0.27
ESC Incubation at 75C At pH 5.5 pH 7.0-grown 7.2 0.53
ESC Incubation at 100C At pH 5.5 pH 7.0-grown 5.2 1.66
ESC Dialysed against broth At pH 5.5 pH 7.0-grown 20.2 0.58
Organisms were grown to log-phase at the stated pH, neutralised, if required, and cell-free filtrates prepared from the cultures grown at pH 7.0
(such filtrates contain ESC) as described for Tables 1 3. These filtrates were treated as indicated (see Section 2 for details of conditions) and
activated at the appropriate pH, followed by neutralisation. Log-phase organisms grown at the stated pH were then incubated with filtrate (1
part culture to one part filtrate) and incubated for 60 min at pH 7.0. After washing with pH 7.0 broth, challenge was at pH 10.5 for 8 min.
N.A. ¼not applicable.
Black plate (145,1) Extracellular sensors and extracellular alarmones 145
Table 5 Killed cultures can induce alkali tolerance and alkali sensitisation
Culture or filtrate used Organisms incubated % survival of organisms after alkali:
for alkali tolerance or with or without
sensitisation induction stated filtrate or culture pH 11.0, 6 min pH 10.5, 8 min
None pH 7.0-grown 0.87 0.13 N.A.
None pH 9.0-grown 23.0 1.95 N.A.
pH 7.0 filtrate pH 7.0-grown 0.33 0.03 N.A.
pH 7.0!pH 9.0 filtrate pH 7.0-grown 17.0 0.82 N.A.
pH 7.0 culture!UV pH 7.0-grown 45.5 2.95 N.A.
pH 7.0 culture,!UV!pH 9.0 pH 7.0-grown 34.9 2.72 N.A.
None pH 7.0-grown N.A. 23.3 1.28
None pH 5.5-grown N.A. 5.07 0.58
pH 7.0 filtrate pH 7.0-grown N.A. 22.1 3.4
pH 7.0!pH 5.5 filtrate pH 7.0-grown N.A. 2.5 0.52
pH 7.0 culture!70C pH 7.0-grown N.A. 19.2 0.75
pH 7.0 culture!70C!pH 5.5 pH 7.0-grown N.A. 5.4 0.14
pH 5.5 culture!70C pH 7.0-grown N.A. 5.66 0.37
For alkali tolerance studies, organisms were grown at pH 7.0. Part of the culture was used to produce filtrates and the other part killed with
UV irradiation (300 sec from a Philips TUV 6W tube at 15cm from the culture). The filtrates and killed cultures were then activated at pH 9.0
if stated, and then midlog-phase organisms grown at the stated pH values were incubated (60min at 37C) with or without filtrate or killed
culture (1 part filtrate or killed culture with 1 part of log-phase culture). After washing the incubated mixtures with pH 7.0 broth, challenge
was at pH 11.0 for 6 min. Similar procedures were used for studying alkali sensitisation by killed cultures, although killing was by heat (70C
for 15 min) and alkali challenge at pH 10.5 for 8 min. N.A. ¼not applicable. Results are given as mean survival percentages SEM.
Black plate (146,1)
3.5 An ESC functions in the acid sensitivity induction
(ASI) process appearing at pH 9.0
Organisms of strain 1157 grown at pH 9.0 or transferred from pH
7.0 to pH 9.0 are highly acid sensitive (Table 6). As previously
an EIC, or possibly more than one EIC, is present
in filtrates from organisms grown at pH 9.0 (Table 6), since such
filtrates, after neutralisation, can induce acid sensitivity if incubated
with pH 7.0-grown cultures at pH 7.0. In contrast, filtrates from pH
7.0-grown cultures cannot induce sensitivity. They do, however,
contain an ESC since these filtrates gain the ability to induce acid
sensitivity if incubated at pH 9.0, without organisms, and neutra-
lised (Table 6). Treatment of the pH 7.0 filtrate with protease, prior
to activation at pH 9.0, reduced the sensitisation by the activated
filtrate to about 60% of that shown with the control (untreated)
filtrate, and RNase had a similar effect.
3.6 Killed cultures can induce ASI
Cultures of strain 1157 grown at pH 7.0 and then subjected to
heating at 65C were able to induce ASI in living cultures if they
were (after killing) activated at pH 9.0, and then neutralised, before
incubation with the living culture at pH 7.0 (Table 7). Also pH 9.0-
grown cultures if neutralised and then killed by exposure to 65C
were able to induce ASI in pH 7.0-grown cultures (Table 7).
Strikingly, the ability to induce ASI of these killed cultures
containing ESC was almost abolished by protease (Table 7).
4. Discussion and future work
Filtrates from pH 7.0-grown strain 1829 cannot induce alkali
sensitisation in organisms at pH 7.0 and accordingly, unlike
filtrates from pH 5.5-grown cultures, do not contain an alkali
sensitisation EIC. These filtrates do, however, contain an EC
since they can be activated when incubated, without organisms,
at pH 5.5 i.e. they contain an EC which is converted to EIC at
mildly acidic pH (Table 4). This EC is formed in activated and non-
activated cultures and is converted, on exposure to the activating
pH, to a component (EIC) which induces the response in pH 7.0-
grown organisms. On this basis, this EC is an acidity detecting
sensor and an ESC. This ESC is a small (dialysable) very heat-
resistant protein. It acts as an acidity sensor, but differs from the
acid tolerance ESC, which is also a protein, in being resistant to
146 Robin J. Rowbury and Margaret Goodson
Black plate (147,1) Extracellular sensors and extracellular alarmones 147
Table 6 An ESC involved in induction of the acid sensitivity appearing at pH 9.0 in E. coli 1157
Filtrate added to
induce sensitisation
of filtrate before
of filtrate
(for 30 min)
Culture incubated
with or without
% survival after acid
for organisms incubated
with or without filtrate
None N.A. N.A. pH 7.0-grown 25.8 1.75
None N.A. N.A. pH 9.0-grown 0.09 0.01
From pH 7.0-grown culture None None pH 7.0-grown 26.2 2.2
From pH 9.0-grown culture None None pH 7.0-grown 1.4 0.17
From pH 7.0-grown culture None At pH 9.0 pH 7.0-grown 6.0 0.17
From pH 7.0-grown culture With protease At pH 9.0 pH 7.0-grown 13.7 0.75
From pH 7.0-grown culture With RNase At pH 9.0 pH 7.0-grown 11.5 1.09
From pH 7.0-grown culture With DNase At pH 9.0 pH 7.0-grown 1.6 0.12
From pH 7.0-grown culture At 75C At pH 9.0 pH 7.0-grown 5.1 0.60
Strain 1157 was grown to log-phase at the stated pH and, where required, filtrates were prepared as described in Section 2, treated as stated
(conditions for treatment are given in Section 2) and activated at the indicated pH. Log-phase cultures of strain 1157 were then mixed with
filtrates (1 part filtrate with 1 part culture) and incubated for 60 min. After washing with pH 7.0 broth, organisms were challenged at pH 3.0
for 7 min; results are shown as % survival SEM values. N.A. ¼not applicable.
Black plate (148,1)
148 Robin J. Rowbury and Margaret Goodson
Table 7 Induction of ASI in strain 1157 by appropriate killed cultures
Culture or filtrate
used for tolerance
Activation for
filtrate or
killed culture
Organisms incubated
with or without
filtrate or killed
% survival ( SEM)
after challenge, for living
culture after incubation
with filtrate or killed
None N.A. pH 7.0-grown 23.6 1.26
None N.A. pH 9.0-grown 0.06 0.03
Filtrate from pH 7.0-grown culture None pH 7.0-grown 22.9 2.6
Filtrate from pH 7.0-grown culture At pH 9.0 pH 7.0-grown 5.8 0.18
pH 7.0-grown 65C-killed culture None pH 7.0-grown 6.5 1.1
pH 7.0-grown 65C-killed culture At pH 9.0 pH 7.0-grown 7.2 0.52
pH 7.0-grown culture, killed at
65C then treated with protease
At pH 9.0 pH 7.0-grown 21.6 3.0
pH 9.0-grown 65C-killed culture N.A. pH 7.0-grown 5.8 0.36
pH 9.0-grown culture, killed at
65C then treated with protease
N.A. pH 7.0-grown 11.6 0.08
Strain 1157 was grown to midlog-phase at pH 7.0 or pH 9.0 and, after neutralisation, cell-free filtrates prepared from the pH 7.0-grown
culture. Filtrates were then activated if appropriate. The neutralised strain 1157 cultures were killed by exposure to 65C for 15 min and then,
when appropriate activated. Filtrates or killed cultures (1 part) were then incubated with pH 7.0-grown midlog phase strain 1157 cultures (1
part) for 60 min at pH 7.0, prior to challenge of mixtures (washed with pH 7.0 broth) at pH 3.0 for 7 min. N.A. ¼not applicable.
Black plate (149,1)
exposure to 100C; the acid tolerance ESC resists 75C, but is
destroyed at 100C
. These two ESCs also differ in size, the acid
tolerance ESC being too large to pass through dialysis tubing. The
alkali tolerance process also involves the functioning of an ESC or
possibly more than one ESC. The evidence for this is that filtrates
from pH 7.0-grown cultures cannot induce alkali tolerance on
incubation with organisms at pH 7.0, but can do so if they are
first incubated (activated) at pH 9.0. By analogy with the system
which switches on acid tolerance, it can be concluded that the pH
7.0 filtrates contain an alkali sensing ESC(s) which is converted to
an EIC at the higher pH. It is possible that there are two ESCyEIC
pairs involved in alkali tolerance induction because the pH 7.0
filtrate is only partially inactivated by protease; a protein ESC and
a non-protein one could, therefore, be present in the pH 7.0
filtrates, with these being converted to a pair of alkali tolerance
inducing EICs by exposure to alkaline pH. The fact that dialysis
leads to loss of the same amount of ESC activity as protease
treatment does (Table 2) would also be in accord with two ESCs,
one being a small protein, removed by dialysis and destroyed by
protease, and the other being a larger non-protein component. The
finding of partial RNase-sensitivity of the ESC-containing prepara-
tion (Table 2) may mean that the second ESC is an RNA. The
results in Table 6 indicate that culture fluids from pH 7.0-grown
cultures of strain 1157 contain an extracellular component which
can be converted to an ASI-inducing EIC at pH 9.0. The fact that
this EC is synthesised under non-stressing conditions and is
converted to an induction component at alkaline pH, in the
absence of organisms, indicates that it has the properties of a
stress sensing component, but is an extracellular sensor, namely an
ESC, which senses alkaline pH and is converted by it to an acid
sensitisation EIC. So far, we have implied that there is a single
ESCyEIC pair, but it is possible that more than one ESC is
involved in the switching on of ASI, since protease treatment of
ESC-containing filtrates only partially inactivates them (Table 6).
Accordingly, ASI may involve both a proteinaceous ESC and a
non-protein ESC, each activated at pH 9.0, with activation produ-
cing conversion to a pair of EICs. Previous studies gave evidence
for two EICs in this system, a protein and a non-protein compo-
, and this, and the present results, would be in accord with
the ASI response consisting of two sensitisation processes, one
involving protein synthesis, the other being protein synthesis
. The finding that RNase partially inactivates the
ESC-containing preparation, suggests the possibility that the two Extracellular sensors and extracellular alarmones 149
Black plate (150,1)
ESCs are a protein and a ribonucleic acid, which could be
converted to corresponding EICs, with these latter components
being able, after entry into the cell, to each interact with appro-
priate regions of DNA to switch on the two ASI sensitisation
processes. The finding (Table 2) that RNase partially inactivated
the alkali tolerance inducing ESC-containing preparations suggests
that ASI and the alkali tolerance response may resemble one
another in each having protein and ribonucleic acid ESCs (they
are, however, quite distinct in the nature of the regulatory
components involved in induction). For two of the stress responses
studied here, the ESCs appear to exist in more than one form, the
form synthesised depending on the cultural conditions during
synthesis. To describe one system in more detail, Table 3 indicates
that the alkali tolerance ESC formed at pH 5.5 differs from that
formed at pH 7.0 in that the former is substantially activated at pH
7.5 whereas the latter is not. Accordingly, if the pH of an acidified
culture rapidly changes to a potentially lethal high pH due
the entry of an alkaline effluent, alkali tolerance can be induced,
during the process, at quite low pH, making it more probable that
the organisms will survive the lethal stress which follows. The EICs
for each system described here can induce the appropriate response
in uninduced cultures and, in view of the fact that all the EICs
studied here are fairly small diffusible components, the appearance
and diffusion of these ECs in the environment can allow cross-talk
between stressed and unstressed cultures with the resulting inter-
cellular communication leading to the unstressed ones being induced
to tolerance prior to being subjected to the stress; such systems
clearly constitute stress early warning systems. The results in Tables
5 and 7 show that appropriate killed cultures appear to induce
alkali tolerance or sensitivity or acid sensitivity in pH 7.0-grown
organisms. Plating of killed cultures on NA showed that all of
those used contained less than 0.01% viable organisms.
Accordingly, the organisms in the mixtures (of killed cultures and
living ones) which show the modified properties are not survivors
of the killing processes used, but are indeed organisms from the
living cultures which have had tolerance or sensitisation properties
induced in them, by the incubation with the killed culture. It seems
highly likely, that as for acid tolerance conferred by killed prepara-
, it is EICs in the killed preparations that confer stress
responses (EICs are resistant to most killing treatments). These
findings may be of public health importance; if food constituents
which have been freed of viable organisms by chemical treatment,
heating, irradiation or filtration (treatments which do not destroy
150 Robin J. Rowbury and Margaret Goodson
Black plate (151,1)
EICs), were subsequently ingested with contaminated food or
water, they could alter the stress tolerance properties of the
contaminating organisms, and this might allow them to go on to
cause disease. Strikingly, one heat-killed ESC-containing prepara-
tion, namely that used (after pH 9.0 activation) for ASI induction,
was completely inactivated by protease (Table 7), suggesting that
the ribonucleic acid ESC is fully inactivated by heat. As noted
above, of particular interest here is that, for two of the processes
considered, there appear to be two pairs of ESCs and EICs, one
pair being proteins and the other pair RNAs. It is important that
these findings are confirmed by taking purified EC isolates and
undertaking further analysis.
E. coli strain 1157 and its tolerance and sensitisation responses
Strain 1157 is unusual in some of its tolerance responses. First, it
fails to induce acid tolerance appreciably in short periods at pH 5.0.
Examination of EC synthesis and functioning have established that
this strain forms the acid tolerance ESC and EIC normally but
shows very slow response to the EIC. Strain 1157 is phoE i.e. lacks
an outer membrane porin, and it seems possible that interaction of
the acid tolerance EIC with the sensitive cell, leading to tolerance
induction, involves PhoE. Several findings are in accord with this.
First, two phoE
transductants show normal acid tolerance at pH
5.0 and second, phosphates, and particularly polyphosphates,
which enter the cell via PhoE, stop acid tolerance induction at
pH 5.0 by normally proficient strains. Involvement of PhoE in acid
tolerance induction by the EIC would not be unexpected, since
there is strong evidence that this porin is involved in numerous
processes relating to acid sensitivity and resistance
. Strikingly,
strain 1157 also fails to induce alkali tolerance normally at pH 9.0,
and the lesion is suppressed in phoE
transductants. There is no
information on alkali tolerance ESCyEIC synthesis or functioning,
but it seems likely that PhoE is implicated in the cellular interaction
with this EIC. Although strain 1157 is deficient in acid and alkali
tolerance responses, several other stress responses are normal,
particularly the ASI response (see above, Section 3.5). Contrary
to powerful contradictions made on theoretical grounds, the outer
membranes (OMs) of most enterobacteria strongly impede proton
passage, this being evidenced by the fact that numerous genetical
alterations and physical or chemical treatments which permeabilise
the OM, but do not alter the cytoplasmic membranes, greatly
enhance acid sensitivity of organisms and the sensitivity to external Extracellular sensors and extracellular alarmones 151
Black plate (152,1)
protons of periplasmic components
. Also, as will become clear
below, at least two inducible acid sensitisation responses, involve
modification or induction of OM porins, again suggesting that the
OM normally impedes proton passage. One of these responses,
namely acid sensitivity induced by L-leucine is normal in strain
1157 (ruling out any functioning of PhoE), but the other response,
acid sensitivity induced by salt, is absent from this strain, possibly
implicating PhoE. We now know that the L-leucine and salt
responses act by enhancing proton passage through novel porins
or normal ones induced to higher levels (see below), and so by
analogy with these, the likelihood is that for ASI, a novel outer
membrane (OM) component, which allows proton passage across
the OM, is induced or is modified, but so far, the nature of the
outer membrane protein responsible has not been identified. Salt-
induced acid sensitivity, in accordance with the absence of this
response from strain 1157, is associated with PhoE induction, a
phoE-lacZ-encoded b-galactosidase being strongly induced by
transfer to salt medium
; as well as being involved in EIC
functioning, this porin allows passage of protons, probably
carried on peptides, and it is this proton passage, increased by
PhoE induction, that leads to sensitisation by salt. Clearly, this
could not be the basis for ASI, as strain 1157 is phoE. The L-
leucine-induced acid sensitisation, which does occur in strain 1157,
involves modification of the OmpA protein, allowing it to function
as a proton pore
. Strikingly, in spite of being similar in size to
the OmpC and OmpF porins, OmpA is not normally a pore,
although Nikaido has defined conditions that allow it to show
weak porin activity. In contrast, the modified OmpAp, present after
L-leucine exposure, shows very substantial proton passage activity.
In accord with L-leucine-induced acid sensitivity using OmpAp for
the extra proton passage needed for acid sensitisation, first, ompA
mutants lacking the protein do not show the L-leucine response
and second, organisms induced for this response show reduced
sensitivity to phages that enter the cell after using the normal
OmpA protein as receptor (suggesting that such organisms have
altered OmpAp). Thus, OmpA must be present for the L-leucine
response. Chloramphenicol-resistant protein synthesis is needed for
this response, so that the novel porin may be formed of OmpAp
and another newly synthesised peptide or protein (probably a
peptide or small protein, as small outer membrane proteins often
shows unusual synthesis characteristics). In contrast, to the proper-
ties of strains totally lacking this protein, many strains having
abnormal OmpAp, including many point mutants are normal in the
152 Robin J. Rowbury and Margaret Goodson
Black plate (153,1)
response, i.e. many ompA mutants, with altered protein show this
response. As might be expected, the only point mutants in ompA
which are aberrant in the response have lesions on the surface of
the protein i.e. where initial interactions between H
(or H
attached to e.g. a carrier peptide) and the protein would be
expected to occur, if the insertion of the OmpA portion of the
modified protein into the OM is similar to insertion of the
unmodified protein in strains growing in the absence of L-
leucine. There are four loops of OmpAp which are normally
close to the outer surface of the OM and these are in the regions
of amino acids 25, 70, 110 and 155. Strikingly, mutations in the
ompA gene leading to the change Gly28 ?Val i.e. in the first surface
loop, Gly70 ?Asp in the second and Val110?Asp in the third did
not prevent L-leucine-induced acid sensitivity; each of these
involves alteration to a neutral amino acid. In contrast, lesions
leading to loss of acidic amino acids, from surface loops, led to a
defective L-leucine response. Thus, Glu68 ?Gly or Glu68 ?Lys,
both changes occurring in the second surface loop, and deletion of
Asp105, occurring in the third loop, abolished sensitisation.
5. Conclusions
Why have extracellular sensors and induction components evolved?
There could be several possible explanations, but two are clear
favourites. First, for systems involving chemical stressors, the
presence of an external sensor allows a. immediate activation and
b. the effect of stressor at its full external concentration, because it
does not need to penetrate to activate the sensor, and, therefore,
would not be diluted before reaching the sensor. Secondly, and far
more critically, for systems involving chemical agents or physical
stresses, the external nature and diffusibility of the EIC that results
from sensor activation, means that this molecule can diffuse to
regions not yet exposed to the stressor, and there give early warning
to unstressed organisms of impending danger, and activate them to
stress tolerance prior to stress exposure. This type of system that
allows organisms to become tolerant without even facing the lethal
agent is unique. Studies where organisms forming EICs were
separated by membranes from unstressed ones
, showed that the
EICs can cross into the unstressed culture and induce it to
tolerance. Strikingly, where organisms unable, for genetical or
physiological reasons, to form EICs were in the second compart-
ment either with or without the stressor, the EICs were able to
diffuse across and induce the defective strains to tolerance i.e. this Extracellular sensors and extracellular alarmones 153
Black plate (154,1)
means that in the environment, if EICs were formed in one region,
they could diffuse to a second region, where genetically or
physiologically defective organisms were present, and there
induce tolerance.
The studies described here on ESCs and EICs were supported by a
Research Grant (to R.J.R.) from the Royal Society and the authors
express their thanks for this award.
1. Russell, A. D. (1984) Potential sites of damage in micro-organisms
exposed to chemical or physical agents. In: M. H. E. Andrew and A. D.
Russell (eds.) The Revival of Injured Microbes, pp. 1 – 18. Academic Press,
New York.
2. Rowbury, R. J., Goodson, M. and Whiting, G. C. (1989) Habituation of
Escherichia coli to acid and alkaline pH and its relevance for bacterial
survival in chemically polluted natural waters. Chem. Indy., 685 – 686.
3. Khazaeli, M. B. and Mitra, R. S. (1981) Cadmium-binding component in
Escherichia coli during accomodation to low levels of this ion. Appl.
Environ. Microbiol., 41, 46– 50.
4. Wolffe, A. P. (1995) The cold-shock response in bacteria. Sci. Prog., 78,
301 – 310.
5. Hardy, K. G. (1975) Colicinogeny and related phenomena. Bacteriol. Rev.,
39, 464– 515.
6. Elliott, E. L. and Colwell, R. R. (1985) Indicator organisms for estuarine
and marine waters. FEMS Microbiol. Revs., 32, 61– 79.
7. Humphrey, T. J., Slater, E., McAlpine, K., Rowbury, R. J. and Gilbert,
R. J. (1995) Salmonella enteritidis PT4 isolates more tolerant of heat, acid
or hydrogen peroxide also survive longer on surfaces. Appl. Environ.
Microbiol., 61, 3161 – 3164.
8. Samson, L. and Cairns, J. (1977) A new pathway for DNA repair in
Escherichia coli.Nature,267, 281– 283.
9. Demple, B. and Halbrook, J. (1983) Inducible repair of oxidative damage
in Escherichia coli.Nature,304, 466– 468.
10. Mackey, B. M. (1984) Lethal and sub-lethal effects of refrigeration,
freezing and freeze-drying on microorganisms. Symp. Soc. Appl. Bacter-
iol., 12, 45– 75.
11. Mackey, B. M. and Derrick, C. M. (1986) Changes in the heat resistance
of Salmonella typhimurium during heating at rising temperatures. Lett.
Appl. Microbiol., 4, 13– 16.
12. Humphrey, T. J., Richardson, N. P., Gawler, A. H. L. and Allen, M. A.
(1991) Heat resistance in Salmonella enteritidis PT4 and the influence of
prior exposure to alkaline conditions. Lett. Appl. Microbiol., 12, 258 – 260.
13. Humphrey, T. J. (1981) The effects of pH and organic matter on the
154 Robin J. Rowbury and Margaret Goodson
Black plate (155,1)
death-rates of salmonellas in chicken scald-tank water. J. Appl. Bacteriol.,
51, 27– 39.
14. Segal, A. W., Geisow, M., Garcia, R., Harper, A. and Miller, R. (1981)
The respiratory burst of phagocytic cells is associated with a rise in
vacuolar pH. Nature,290, 406– 409.
15. Lee, A. and Gemmell, E. (1972) Changes in the mouse intestinal micro-
flora during weaning: role of volatile fatty acids. Infect. Immunol., 5,17.
16. Jensen, M. S. and Bainton, D. F. (1973) Temporal changes in pH within
the phagocytic vacuole of the polymorphonuclear neutrophilic leukocyte.
J. Cell. Biol., 56, 379– 388.
17. Leyer, G. J. and Johnson, E. A. (1993) Acid adaptation induces cross-
protection against environmental stresses in Salmonella typhimurium.
Appl. Environ. Microbiol., 59, 1842– 1847.
18. Goodson, M. and Rowbury, R. J. (1990) Habituation to alkali and
increased UV-resistance in DNA repair-proficient and deficient strains
of Escherichia coli grown at pH 9.0. Lett. Appl. Microbiol., 11, 123 – 125.
19. Rowbury, R. J., Goodson, M. and Humphrey, T. J. (1993) Novel acid
sensitivity induced in Escherichia coli at alkaline pH. Lett. Appl. Micro-
biol., 16, 223– 227.
20. Rowbury, R. J. and Hussain, N. H. (1996) Exposure of Escherichia coli to
acid habituation conditions sensitizes it to alkaline stress. Lett. Appl.
Microbiol., 22, 57– 61.
21. Neidhardt, F. C., Ingraham, J. L. and Schaechter, M. (1990) Physiology of
the Bacterial Cell. Sinauer Associates Inc: Sunderland, Mass.
22. Igo, M. M. and Silhavy, T. J. (1988) EnvZ, a transmembrane sensor of
Escherichia coli is phosphorylated in vitro.J. Bacteriol., 170, 5971 –
23. Repaske, D. R. and Adler, J. (1981) Changes in intracellular pH of
Escherichia coli mediate the chemotactic response to certain attractants
and repellents. J. Bacteriol., 145, 1196– 1208.
24. Rowbury, R. J. (1999) Acid tolerance induced by metabolites and secreted
proteins and how tolerance can be counteracted. Novartis Found. Symp.,
221, 93– 111.
25. Rowbury, R. J. and Goodson, M. (1999) An extracellular acid stress-
sensing protein needed for acid tolerance induction in Escherichia coli.
FEMS Microbiol. Lett., 174, 49– 55.
26. Rowbury, R. J. and Goodson, M. (1999) An extracellular stress-sensing
protein is activated by heat and u.v. irradiation as well as by mild acidity,
the activation producing an acid tolerance-inducing protein. Lett. Appl.
Microbiol., 29, 10– 14.
27. Lazim, Z. and Rowbury, R. J. (2000) An extracellular sensor and an
extracellular induction component are required for alkali induction of
alkyl hydroperoxide tolerance in Escherichia coli.J. Appl. Microbiol., 89,
651 – 656.
28. Rowbury, R. J. (2001) Extracellular sensing components and extracellular
induction component alarmones give early warning against stress in
Escherichia coli.Adv. Microbiol. Physiol., 44, 215 – 257.
29. Rowbury, R. J. (2000) Killed cultures of Escherichia coli can protect living
organisms from acid stress. Microbiol., 146, 1759 – 1760. Extracellular sensors and extracellular alarmones 155
Black plate (156,1)
30. Rowbury, R. J., Lazim, Z. and Goodson, M. (1996) Regulatory aspects of
alkali tolerance induction in Escherichia coli.Lett. Appl. Microbiol., 22,
429 – 432.
31. Rowbury, R. J. and Hussain, N. H. (1998) The role of regulatory gene
products in alkali sensitisation by extracellular medium components in
Escherichia coli.Lett. Appl. Microbiol., 27, 193– 197.
32. Rowbury, R. J., Goodson, M. and Wallace, A. D. (1992) The PhoE porin
and transmission of the chemical stimulus for induction of acid resistance
(acid habituation) in Escherichia coli.J. Appl. Bacteriol., 72, 233– 243.
33. Rowbury, R. J. and Goodson, M. (1993) PhoE porin of Escherichia coli
and phosphate reversal of acid damage and killing and of acid induction
of the CadA gene product. J. Appl. Bacteriol., 74, 652 – 661.
34. Rowbury, R. J. (2004) Enterobacterial responses to external protons,
including responses that involve early warning against stress and the
functioning of extracellular pheromones, alarmones and varisensors. Sci.
Prog., 87, 193– 225.
35. Rowbury, R. J. (2003) Physiology and molecular basis of stress adapta-
tion, with particular reference to the subversion of stress adaptation and to
the involvement of extracellular components in adaptation. In: A. E.
Yousef and V. K. Ahmed (eds.) Microbial Stress Adaptation and Food
Safety, pp. 247 – 302. CRC Press. Boca Raton, Florida.
36. Rowbury, R. J. (2005) Stress responses of foodborne pathogens, with
specific reference to the switching-on of such responses. In: P. M.
Fratamico, A. K. Bhunia and J. L. Smith (eds.) Foodborne Pathogens:
Microbiology and Molecular Biology, pp. 77 – 97. Caister Academic Press.
Wymondham, UK.
156 Robin J. Rowbury and Margaret Goodson
Black plate (157,1)
The molecular basis of lactose
A staggering 4000 million people cannot digest lactose, the sugar in milk,
properly. All mammals, apart from white Northern Europeans and few tribes
in Africa and Asia, lose most of their lactase, the enzyme that cleaves lactose
into galactose and glucose, after weaning. Lactose intolerance causes gut and
a range of systemic symptoms, though the threshold to lactose varies
considerably between ethnic groups and individuals within a group. The
molecular basis of inherited hypolactasia has yet to be identified, though two
polymorphisms in the introns of a helicase upstream from the lactase gene
correlate closely with hypolactasia, and thus lactose intolerance. The
symptoms of lactose intolerance are caused by gases and toxins produced
by anaerobic bacteria in the large intestine. Bacterial toxins may play a key
role in several other diseases, such as diabetes, rheumatoid arthritis, multiple
sclerosis and some cancers. The problem of lactose intolerance has been
exacerbated because of the addition of products containing lactose to various
foods and drinks without being on the label. Lactose intolerance fits exactly
the illness that Charles Darwin suffered from for over 40 years, and yet was
never diagnosed. Darwin missed something else – the key to our own
evolution – the Rubicon some 300 million years ago that produced lactose
and lactase in sufficient amounts to be susceptible to natural selection.
Keywords:lactose, lactase, lactose intolerance, milk, hypolactasia,
evolution, Darwin, bacterial toxins
Department of Medical Biochemistry and Immunology, Wales College of Medicine,
Cardiff University, Cardiff, CF14 XN, UK
Department of Medical Biochemistry and Immunology, Cardiff and Vale NHS Trust,
Llandough Hospital, Llandough, Penarth, Vale of Glamorgan, CF64 2XX, UK
Science Progress (2005), 88(3), 157–202 157
Black plate (158,1)
Many individuals can ingest milk without any problem. However,
many thousands of others suffer debilitating symptoms from just
one glass of milk or less because they are intolerant to lactose, the
sugar in milk. If we are to help these people, whose lives are misery
because of their sensitivity to lactose, we need to answer five key
.What is the precise cause of lactose intolerance?
.How does lactose produce gases and toxins in the gut?
.How do these toxins cause the wide range of gut and non-gut
symptoms that make the lives of people with lactose intolerance
a misery?
.What was the evolutionary origin of this biochemical system,
unique to mammals?
.How has it evolved over the following 300 million years to
influence our diet and health in the 21st century?
An unusual case
A few years ago we discovered a 53 year old woman whose life was
a misery because of severe irritable bowel syndrome (IBS), diar-
rhoea, nausea and sickness, as well as skin rashes, breathing
problems, muscle and joint pain, and lack of concentration. She
had suffered these since childhood. But they were now so severe
that she thought she had Alzheimer’s disease. Her doctor told her
she had eczema, asthma and osteo-arthritis. She was awaiting a
knee replacement operation, and was on a range of drugs – skin
creams, antihistamines, asthma inhalers, antibiotics, anti-diar-
rhoeals and strong pain relief. She was surprised when we
decided to investigate her for lactose intolerance 50 g of oral
lactose, followed by an analysis of breath for hydrogen gas. But
this was negative. According to the text books, this lady did not
have lactose intolerance, as she had many non-gut symptoms and
did not produce hydrogen when she ingested lactose. But she did
become ill during the lactose test, recording gut and non-gut
(systemic) symptoms several hours after taking the lactose – gut
pain, nausea and vomiting, headache, light headedness, feeling
drunk, heart palpitations, and joint and muscle pain. These
remained severe for three days. We advised her to remove all
lactose from her diet for one month, involving avoidance of
‘dairy’ products, and foods and drinks where lactose can be
158 Anthony K. Campbell, Jonathan P. Waud and Stephanie B. Matthews
Black plate (159,1)
present without being clear on the label. Within one month she
described her skin as ‘wonderful’. Her asthma and sinusitis had
gone, and her joints were much improved. She no longer needed
any medication, and was taken off the list for a knee replacement.
Similar dramatic stories have been repeated among the 700 patients
we have now diagnosed as having ‘systemic’ lactose intolerance
What is lactose intolerance?
The fact that many southern Europeans become ill after drinking
milk was first described by Hippocrates. But it took 2000 years to
discover that this was caused solely by a biochemical intolerance to
the sugar in milk. Lactose intolerance was first identified in the
early years of the 20th century
. However, it was not until the
1960s that the biochemical basis of lactose intolerance, and its
ethnic distribution, were properly defined
The disaccharide lactose, 4-O-b-D-galactopyranosyl-D-gluco-
pyranose (Figure 1), is found widely in Nature attached to
polysaccharides, glycoproteins and glycolipids. The latter involve
gluco- and galacto-ceramides, and lipids involved in the vesicles of
endocytosis, e.g. lactose attached to the sialic acid N-acetyl
glucosamine as neuramin lactose (Figure 2) found in small quan-
tities in milk. Large amounts of free lactose are only found
naturally in mammalian milk, where it can exist in the interconver-
table aor bforms (Figure 1). alactose is the principle form in milk,
and that supplied in a Pharmacy, and used in the lactose test.
Lactose dissolves in water up to about 1.5 M in boiling water. But it
is not as soluble as glucose, fructose or sucrose. Molar solutions of
lactose come out of solution when frozen, unlike glucose and
sucrose. Milk does not taste sweet because alactose has only a
faintly sweet taste, the bform being slightly sweeter. Both forms
rotate the plane of polarised light (aþ92:68, and bþ348at 20C).
Warming either form leads to an equilibrium value of the aþb
forms with an optical rotation coefficient ½a20
Dof 52.3.
All mammalian milk, apart from that of the Pinnepedia (sea lions
and walruses), contains 40 – 75 g lactose per litre, depending on
species, providing 40% of the energy needs of a suckling infant. Yet
some two-thirds of the world’s population cannot digest lactose
properly (Table 1). Each of us has a different threshold to lactose.
Many white Northern Europeans can drink 1 – 2 glasses (250 –
500 ml) of milk with no adverse effects, while others are so sensitive
to lactose that just 10 20 ml in a cup of tea can make them ill. This
is because: The molecular basis of lactose intolerance 159
Black plate (160,1)
160 Anthony K. Campbell, Jonathan P. Waud and Stephanie B. Matthews
Fig. 1. The chemical structure of some sugars. The aform of lactose has the hydroxyl at position 1 on the glucose up, rather than down as in the b
Black plate (161,1)
1. A person labelled as lactose intolerant has a low level of lactase,
the enzyme in the small intestine that cleaves lactose into glucose
and galactose, which can then be absorbed.
2. Bacteria in the large intestine convert any lactose undigested in
the small intestine into gases and toxins.
3. The tissues are sensitive to the bacteria toxins, after they have
been absorbed into the rest of the body.
We are dealing with a biochemical intolerance, and not an allergy,
though lactose intolerance can exacerbate allergic symptoms
allergy involves an immune response to a foreign protein, resulting
in a reaction with antibodies IgE or IgG. These antibodies bind the
allergen, and then the antigen-antibody complexes activate cells in
the immune system – lymphocytes to generate more antibodies,
phagocytes to release oxygen metabolites and proteases, and
importantly, mast cells to release histamine. These substances
then cause contraction of smooth muscle and inflammation, with The molecular basis of lactose intolerance 161
Fig. 2. The chemical structure of some other substances relevant to lactose
Black plate (162,1)
resulting breathing difficulties, skin itching and rashes. In its
severest form a patient may suffer anaphylactic shock that can be
lethal. In contrast, an intolerance is a biochemical defect that
prevents the normal metabolism of a specific substance. Most
commonly, such biochemical intolerances are to a carbohydrate,
amino acid or other small organic molecule
. Lactose intolerance
is caused by an impaired capacity to digest lactose properly, and
thus a reduced capacity to absorb into the body its two constitutive
sugars, galactose and glucose. To understand fully lactose intoler-
ance five questions need to be answered:
1. How is lactose normally digested, and what mechanisms can
prevents this occurring?
2. When lactose is not digested normally, what happens to it?
3. What are the symptoms that result from lactose not being
digested normally?
4. What is the molecular basis of these symptoms, and what causes
someone to cross the Rubicon and feel ill?
5. What is the evolutionary significance of lactose?
162 Anthony K. Campbell, Jonathan P. Waud and Stephanie B. Matthews
Table 1 Different ethnic groups with low lactase and likely lactose intolerance
Ethnic group (adult, unless stated)
% with low
lactase and potential
lactose intolerance
Chinese 490%
Japanese 490%
Indian and other Asian groups 480%
Aboriginal Australian 480%
Black African 475%
American Red Indian 470%
Eskimo 470%
South American (total adults) 450%
Mexican 450%
West Indian 450%
Spanish 440%
Italian 440%
Greek 440%
Mid European (e.g. Hungarian and gypsy) 440%
American (total adults) 30%
Finnish 20%
White Northern European 10%
White Australian 10%
Children under 2 years old (any ethnic group) 0 – 20%
Children between 2 and 10 years old 0 – 40%
Patients with IBS 450%
These numbers are very approximate.
Black plate (163,1)
The discovery of lactose and lactase
Lactose was discovered in milk in the 17th century
. But it took a
further 300 years before lactose was synthesised in the laboratory
and thus its precise chemical structure determined. The fact that
lactose can induce diarrhoea was reported over 100 years ago
was then shown that animal and human intestine contained an
enzyme, lactase, that could cleave lactose into its two constituent
. Consistent with the fact that sea lion milk has no lactose,
the intestine of these animals seemed devoid of lactase
Free lactose is synthesised from UDP-galactose and glucose in
the mammary gland. Lactose synthase has two sub-units, galac-
tosyl transferase and a protein modifier, a-lactalbumin. Galactosyl
transferase normally catalyses the formation of N-acetyl lactosa-
mine, on glycoproteins:
UDP galactose þNacetyl glucosamine ?
?UDP þNacetyl lactosamine ð1Þ
However, when the modifier sub-unit, a-lactalbumin, binds to
galactosyl transferase the resulting complex changes its specificity
to become a lactose synthase, transferring galactose to glucose
rather than N-acetyl glucosamine:
UDP galactose þglucose ?UDP þlactose ð2Þ
Galactosyl transferase is found in most tissues, but lactose synthase
is found only in the mammary gland, where, in pregnancy, its gene
switches on. At birth, the hormone prolactin then induces the
modifier sub-unit a-lactalbumin, so that the breast can produce
lactose in the milk for the new born baby.
The biochemistry of lactase
Lactase is a special type of b-galactosidase. There are three b-
galactosidases found in human tissues:
1. Specific lactase on the apical surface of the enterocytes in the
brush border villi, facing outwards, having a pH optimum of
about 6. It breaks lactose (C
) into D( þ)-galactose and
D( þ)-glucose, and is found only in the small intestine in
2. b-galactosidase in the cytosol of cells, with a similar pH
optimum but which may not hydrolyse lactose.
3. b-galactosidase in the lysosomes, with a pH optimum of 4.5. The molecular basis of lactose intolerance 163
Black plate (164,1)
In bacteria, the classic lac operon codes not only for a b-
galactosidase but also for a permease, necessary if the bacteria
are to take up lactose and access cytosolic b-galactosidase. It is not
clear whether eukaryotic cells have such an active lactose permease
in the plasma membrane. Yeast metabolises external lactose
. Thus, lactose can be added to drinks, and even beers
and lagers, without generating unnecessary CO
. In most eukaryo-
tic cells, the b-galactosidases see only b-galactosides generated from
internal metabolism, such as those attached to lipids and proteins.
Krabb’s disease (globoid cell leukodystrophy) is caused by a
deficiency in intracellular galactosylcerebrosidase. But it is loss of
intestinal lactase that is responsible for lactose intolerance. It is
found first in the duodenum (5 – 6 cm long), reaches a peak in the
jejunum (2.5 m long), and decreases gradually down the ileum (4 –
5 m long). There are no significant amounts normally found in the
large intestine, first shown over 100 years ago
Intestinal lactase is a unique enzyme, since it has two active sites
within one polypeptide chain. One hydrolyses lactose, while the
other was identified originally by its ability to hydrolyse an aryl
glycoside called phlorizin (Figure 2), discovered in apple bark,
being hydrolysed to glucose and phloretin (Figure 2), a diabeto-
genic substance. Phlorizin is an aryl a-glucoside linked to phloretin
(Figure 2), and was originally discovered as an inhibitor of the
glucose uptake mechanism in the small intestine SGLT1. Phlorizin
is a competitive inhibitor of the lactose site. But lactose does not
appear to inhibit the phlorizin site. Small intestinal lactase is
competitively inhibited by a number of other substances, including
the common buffer Tris, and colchicine (Figure 2) that binds to
tubulin in microtubules. Unlike the acid pH b-galactosidase in
lysosomes, intestinal lactase is not blocked by SH reactive reagents
such as p-chloro-mercurobenzoate (PCMB).
The full name is therefore lactase-phlorizin hydrolase with two
enzyme commission (EC) numbers – EC for its phlorizin
hydrolase (LPH) activity and EC for its b-galactosidase
activity. Care should be taken to use the correct EC numbers, as
there are some publications that have used the incorrect numbers
with the bacterial b-galactosidase number, EC Small
intestinal lactase has no amino acid sequence similarities to the b
galactosidase in bacteria. It is also different from enzyme supple-
ments sold as ‘lactase’ in health food shops, and the other types of
bgalactosidases found in eukaryotic cells.
The natural substrates for the phlorizin site are cerebrosides
(Figure 2) – glycolipids made up of a hexose sugar, usually
164 Anthony K. Campbell, Jonathan P. Waud and Stephanie B. Matthews
Black plate (165,1)
galactose, linked by a blink to sphingosine with a fatty acid
attached. The non-sugar moeity is known as a ceramide. This
enzyme activity is thus really a glycosyl ceramidase. This explains
why we need to keep some lactose after weaning. Hydrolysis of
cerebrosides provides sphingosine, particularly important in the
membranes of the brain. Lactase, thus, has a number of enzymatic
activities in addition to hydrolysing lactose, including a range of b-
glycosides (phlorizin, glycolsyl and aryl b-ceramides, cellulose,
cellobiose, cellotriose, cellotetraose), b-glucans found in the cell
walls of plants and fungi we eat, e.g. laminaribose and to a lesser
extent gentiobiose
, and b-galactosides. It also hydrolyses o- and
m- nitro-phenyl b-glycosides, useful as artificial substrates in
assays. The hydolysis of flavonoid glycosides and pyridoxine-50-
may be important sources of flavonoids and
vitamin B6, forming pyridoxal phosphate used in several energy
system enzymes such as phosphorylase, and amino-transferases.
Lactose is restricted to the milk of terrestrial mammals, but
glycosyl ceramides are present in the diet of all vertebrates.
Lactase is not a particularly powerful enzyme, there being
variations in the maximum enzyme activity (V
) when saturated
with substrate, and the affinity for the substrate (K
) for lactase
and phlorizin hydrolase activity between species
. The ratio of
activities of the lactaseyphlorizin-glycosyl ceramidase activities also
varies between species, from 40 or 35 :1 in rats or monkeys to 5 :1
in humans. This means that rats and monkeys are better at
digesting cerebrosides than humans. Human lactase has a K
lactose of about 20 mM, compared with pig at 5 mM, consistent
with human milk having the higher lactose concentration. The Km
for phlorizin is 0.4 mM in the human and pig enzymes, but
530 mM in pig. Human lactase has a moderate V
of 20 Uymg
pure protein (1 U ¼1mmol min
). If one assumes a molecular
weight of about 150,000, then this gives a turnover number of
about 50 s
. This compares with 600,000 s
for carbonic anhy-
drase, 1000 s
for lactate dehydrogenase and 1 s
for firefly
luciferase. This has important implications for the evolution of
such enzymes. At this stage in their evolution they can simply be
considered as ‘solvent cages’. Natural selection has yet to force
improvement of biochemical properties through covalent and other
interactions with their substrates. But the V
and K
sufficient for lactase to be maximally active with a lactose concen-
tration in cow’s milk of 130 mM, and in human milk of 190 mM.
From a turnover number of 50 s
, it is possible to estimate a total
lactase activity in the entire human small intestine of 2500 U
.It The molecular basis of lactose intolerance 165
Black plate (166,1)
would thus take less than 15 min to digest all the lactose (33 mmole)
in a 250 ml glass of milk. But, in someone who is severely
hypolactasic, with a total lactase level of just 250 U (i.e. 10%),
then it would take over 2 h to digest all this lactose. By this time it
has reached the bacteria in the large intestine. The pH optimum for
lactase is about 6, with little activity below pH 3. So it would be
inactive in the stomach, where the pH is 1 – 3. However the pH rises
in the duodenum to 6 – 6.5, and then in the jejunum and ileum to
pH 7 – 8 as the food gets further away from the stomach. Since the
pH activity curve of lactase is skewed towards alkaline pH, at pH
8 9 lactase still retains 50% of it maximal activity at pH 6, suitable
for full activity throughout the small intestine, becoming more
alkaline as food moves from the stomach to duodenum and then
the ileum. The pH of the large intestine is 5.5 7.
Cellulose is the major polysaccharide in all plant cell walls, made
of long chains of 1 – 4 blinked glucoses, unlike starch where the
glucoses are linked by 1 – 4 and 1 – 6 abonds. a– amylase cannot
hydrolyse cellulose. Ruminants have bacteria in their multiple
stomachs to achieve this efficiently. However lactase can hydrolyse
cellulose, and the di-, tri- and tetra- saccharides
, its initial
degradation products – cellobiose, cellotriose and cellotetraose
(Figure 2). In monkeys
, the specific activity ratio for lacto-
se :cellulase is 6 :1, compared with 40 :1 for phlorizin. So lactase
could be more active in hydrolysing products from cellulose than
ceramides. Lactase does not hydrolyse the laxative lactulose (4-O-
b-D galactopyranosyl-D-fructose; Figure 2), slightly sweeter than
lactose. Lactulose can be used to measure gut transit time, when it
generates gas in the large intestine and induces diarrhoea.
Although intestinal lactase has no sequence homology to the b-
galactosidase in E.coli, a comparison of amino acid sequences,
using the software programme BLAST, of human lactase against
the Genbank database, including genomes, identifies over 1800
proteins with some sequence similarity. However, the only major
sequence similarities are with other intestinal lactases – rabbit, rat,
mouse, cow, dog and pig, either from cloning or predicted from the
genome sequence. Human lactase is 83% identical to that in rabbit,
and 77% identical to rat. But rat is only 75% identical to rabbit.
These produce a score of 43000 bits. Other proteins only have ‘bit’
scores of 600 or less. These include the Klotho precursor, a range of
cytosolic eukaryotic and prokaryotic b-glucosidases, gentobiase,
and myrosinase. BLAST and CLUSTAL data suggest that the
active site for b-glucosidases and b-galactosidases has arisen several
times independently in evolution.
166 Anthony K. Campbell, Jonathan P. Waud and Stephanie B. Matthews
Black plate (167,1)
There are four domains in the initial full lactase prosequence
translated from mRNA, designate I, II, III and IV. Domains I and
II are lost in ER cleavage of the N-terminus. Site directed
mutagenesis and affinity labelling identified domain III as the
phlorizinycerebroside hydrolase, and domain IV is the lactase
. The key amino acids are two glutamates, at position 1273
for human phlorizin hydrolase and 1749 for human lactase. The
same glutamates have been identified in rabbit and rat, but at
slightly different numbered sites, because of the different lengths of
the full sequence. In contrast, the negative amino acid at the active
centre of sucrase appears to be an aspartate. Sucrase-isomaltase
can be isolated as one enzyme complex. However, unlike lactase,
the two enzymatic activities are on different polypeptide chains.
Lactase has no sucrase, isomaltase or amylase activity.
The molecular biology of lactase
Human lactase is located on the long arm of chromosome 2 (2p21q).
The 55 kb DNA sequence contains 17 exons, and lies within a 70 kb
sequence containing regulatory response elements
(Figure 3). It
is on the reverse strand. In humans regulation involves both
transcriptional and post-transcriptional mechanisms, transcrip-
tional regulation controlling appearance of lactase in the foetus
just before birth, and its loss on weaning. The developmental
element responsible for the large increase in lactase just prior to
birth is cis acting, CE-LPH
. In spite of extensive searching, no
mechanism causing hypolactasia after weaning has been identified.
The main reason for this is that potential mechanisms have focussed
on regulation of the lactase response element itself, rather than the
development or survival of cells expressing lactase. Using luciferase
reporters, yeast hybrids, gel shift assays with binding of putative
transcription factors, specific antibodies, mutants, and co-transfec-
tion of particular transcription factors, the main lactase response
element in humans, pig and rat has been located within a 1 kb
stretch immediately upstream from the gene, with four key regions
at 894 to 798, 227 to 142, 299 to 227 and 142 to 17 in
the pig, and potential regulation by the caudal homeodomain
transcription factor cdx2
, HNF1a, HIF (hypoxia-inducible
factor), HOXC11, FREAC, and GATA transcription factors 4, 5
and 6
, known to be gut and stomach homeodomain factors.
Mutations within the 1 kb that prevented transcription factor
binding identified the minimal promoter being – 200 to – 17 from
the translation start site, with two cdx2 sites, one HNF1asite, and The molecular basis of lactose intolerance 167
Black plate (168,1)
168 Anthony K. Campbell, Jonathan P. Waud and Stephanie B. Matthews
Fig. 3. The lactase gene.
Black plate (169,1)
the GATA site being at – 100 to –73. There are homologies between
the human, pig and rat lactase promoters
, though the human
promoter has two tandem alu elements within it. But none can be
linked causally to lactase non-persistence that starts on weaning.
This loss is highly specific for lactase, and does not occur with other
disaccharidases such as sucrase. Lactose does not regulate the
lactase promoter, unlike the induction of b-galactosidase (EC in bacteria.
In order to reach the plasma membrane, lactase undergoes
considerable post-translational modification, involving glycosyla-
tion and proteolytic cleavage
(Figure 4). Human lactase is
synthesised as a pre-proprotein of 1927 amino acids
(1926 in
rabbits) (Figure 4). But, the mature protein at the apical plasma
membrane of the human enterocyte is only 1059 amino acids
(1060 in rabbits). This consists of 1014 amino acids at the N-
terminus facing the gut lumen with a terminal A869, a single
membrane spanning domain of 19 amino acids, and a short C-
terminus of just 26 amino acids facing the cytosol (25 in rabbits).
As the lactase is synthesised from mRNA on the ribosome, the N-
terminal signal peptide translocates it into the endoplasmic
reticulum (ER), where the signal peptide of 19 amino acids is
cleaved. Once inside the ER, the 1908 amino acid polypeptide is
cleaved, almost in two. But, although the final protein in the
membrane has an N-terminal A869, this is not the cleavage site in
the ER. Mutation of R868A does not prevent proteolytic cleavage
and processing of lactase in caco-2 cells
. We have shown, using
a genetically engineered ‘Rainbow’ protein
that no cleavage
occurred around R868 at this site in caco-2 cells.
Two proteolysis steps are required to produce the final 1059
amino acid protein in the plasma membrane. The first cleavage
occurs between R734 and L735 via a furin-like protease, though
the role of furin is not fully established. This occurs through the
trans Golgi network, mutant R734yL735 retaining prolactase in
the ER
, and processing being inhibited by monensin and
brefeldin A. Transportation on microtubules is also needed,
since colchicine inhibits formation of the mature enzyme,
causing precursor accumulation. After cleavage, lactase is glyco-
sylated (15 N-linked predicted) and transported to the plasma
membrane, where gut lumen trypsin trims the protein to the final
1059 peptide with an N-terminal alanine. The cleaved N-terminal
866 (847 þ19) amino acids contain domains I and II (87-172 and
363-848), with sequence similarity to the active site domains III
and IV, 883-1365 and 1370-1841. But the N-terminus does not The molecular basis of lactose intolerance 169
Black plate (170,1)
170 Anthony K. Campbell, Jonathan P. Waud and Stephanie B. Matthews
Fig. 4. Lactase-phlorizin hydrolase (EC 3.1. 2.62 and 108), with cleavage and other sites marked.
Black plate (171,1)
appear to have any b-galactosidase activity. Rather it is a
chaperone, helping folding of the final protein, and its successful
trafficking to the plasma membrane
. The final lactase enzyme
monomer has a molecular weight of about 145 kDa, some 24 kDa
greater than that estimated from the amino-acid sequence alone
(121 kDa), consistent with its heavy N-linked sites via asparagines
and O-linked glycosylation via serines and threonines. These are
necessary for efficient folding and activity of the enzyme on the
cell surface. The final dimer has a measured molecular weight of
about 320 kDa.
There are thus are five ways to reduce intestinal lactase activity:
1. Reduced transcription of the lactase gene and splicing of the
mRNA product.
2. Reduced translation of the mRNA.
3. Impairment of enzyme processing in the ER-Golgi, through
protein mal-folding, improper proteolysis or glycosylation, or
reduced protein transport to the plasma membrane.
4. Inhibition of enzymatic activity by substances in the gut lumen.
5. Loss of ‘lactase’ expressing cells through microbial or viral
damage, mechanical loss, or apoptosis.
Loss of lactase expressing cells is the major mechanism respon-
sible for inherited hypolactasia, and with loss in coeliac disease.
There has been some debate about the role of transcriptional
versus post transcriptional regulation of lactase. Too much
emphasis has been placed on these as the explanation for
lactase persistenceynon-persistence. Enterocytes expressing
lactase in the villi of the small intestine exhibit a patchy appear-
, showing that they are generated by a non-clonal
Eating lactose gradually over several weeks or months may
increase the threshold to lactose, before becoming ill. This is
not caused by specific induction of intestinal lactase. Plimmer
in 1906 showed that lactase was not induced by its substrate
lactose. Any apparent dietary induction of lactase is caused by
general intestinal hypertrophy. The inability of lactose to affect
the level of intestinal lactase in mammals is in striking
contrast to the famous induction by lactose of the lac
operon in E.coli, and intestinal sucrase and isomaltase, which
do not decrease after weaning and are induced by ingestion of
their substrates. Changes in lactose sensitivity reported by
patients are most likely caused by changes in microflora in
the large intestine. The molecular basis of lactose intolerance 171
Black plate (172,1)
Hypolactasia versus lactose intolerance
Most of the world’s adult population are ‘hypolactasic’ (Table 1),
compared with a suckling infant. They have a low lactase, and are
thus lactase non-persistent. There are three mechanisms that can
cause this:
1. Congenital loss. This is very rare, though the genetic defect is
found particularly in the Finnish population
, and appears to
be complete loss of lactase. Until recently, the mechanism of
such alactasia was unknown. Initial studies suggested it maps 50
to the lactase gene, i.e. towards the MCM gene, so may be a
regulatory defect rather than a mutation within the lactase gene
itself. However, mutations have now been identified in the
lactase gene itself, causative of congenital lactase deficiency
Characterisation of five mutations in the coding region of the
lactase gene have shown 84% were homozygous for a nonsense
mutation, T4170A (Y1390X), designated ‘Fin (Major)’. Four
other rare mutations included two that result in a frameshift and
early truncation at S1666fsX1722 and S218fsX224, and two
point mutations that result in substitutions Q268H and
G1363S of the 1927aa polypeptide. All four lead to a protein
structure with inactive enzyme.
2. Inherited loss on weaning. This is the norm in all mammals,
apart from white Northern Europeans and some other ethnic
groups (Table 1).
3. Secondary loss. This can occur as a result of intestinal bacterial,
viral or protozoan infections. These include rotavirus, the pro-
tozoan Giardia and gut trypanosomes. Endocrine control
through sex and thyroid hormones
, and ageing, also may
affect levels of lactase in the small intestine.
It is important to distinguish these when treating someone clini-
cally, or when investigating the genetics. Only secondary loss of
lactase is potentially reversible, and thus treatable. Lactase can also
be reduced in a number of other conditions of food intolerance
such as coeliac disease
In all eukaryotic cells the endoplasmic reticulum (ER) has a
signalling system that communicates to the cytosol, plasma
membrane, and the nucleus
. This system determines whether a
cell fires a Ca
signal to switch on an end response, traverses the
cell through its division cycle, whether the cell defends itself against
stresses such as the generation of large amounts of unfolded protein
in the ER, or dies by apoptosis. Stress to the ER will lead to a
172 Anthony K. Campbell, Jonathan P. Waud and Stephanie B. Matthews
Black plate (173,1)
reduction in the lactase reaching the plasma membrane
. This is
likely to be particularly relevant to loss of lactase in gut infections.
The genetics of lactase
The genetics of lactose intolerance have been studied extensively in
white European populations and non-white populations
throughout the world
. The presentation of the genetics is
not very clear. Most geneticists argue that lactase persistence is an
autosomal dominant trait, whereas non-persistence is recessive.
People who are homozygous for lactase persistence retain high
levels of lactase into adulthood. Those who are homozygous for
lactase non-persistence have low levels of lactase in adulthood.
Adults who are heterozygous have intermediate lactase levels. But
clinically, these distinctions are not so clear. There is a huge
variation between individual phenotype, both in hypolactasia and
the threshold for lactose intolerance. Some patients with most
severe symptoms are heterozygote for certain genetic markers
linked closely to lactase non-persistence. The phenotype is not
‘all or none’.
A problem is how to assess hypolactasia. This is usually done
from a small biopsy. The small intestine is an incredible absorbing
machine, folded over and over again to compact it into the
peritoneal cavity. Fully opened out, as a single cell layer, estimates
of the surface area of the small intestine vary from the size of a
tennis court to half the size of the football pitch at Old Trafford!
This poses problems when interpreting measurements from just a
single biopsy. It is like taking a blade of grass from Cardiff’s
Millennium Stadium to discover whether the whole pitch is fit to
play the Cup Final. It is the overall level of lactase in the entire
small intestine, and the efficiency of the sugar transporter SGLT1,
that determine whether all the lactose is first cleaved and then the
resulting glucose and galactose fully absorbed in the small intestine,
before having a chance to reach the bacteria in the large intestine.
Intestinal lactase levels are low in the foetus, unlike sucrase.
Lactase only appears in the foetus a few days before birth, reaching
a peak some 3 days afterwards, just right for the baby to receive
lactose from the mother’s milk. After weaning, some 6 – 12 months
later, lactase begins to decline. The rate of this decline varies
considerably between ethnic groups
. In Chinese and
Japanese, lactase decreases rapidly after 2 – 3 years of age, reaching
its nadir by the age of 5 – 10 years. In Asians the rate of decline is
slightly slower, but still this group have lost some 75% of their The molecular basis of lactose intolerance 173
Black plate (174,1)
lactase by their teens. In contrast, in the 8 – 10% of white Northern
Europeans who lose lactase after weaning the rate of decline is
slower, lactase not reaching its nadir until almost adulthood. The
ultimate level of lactase also varies between ethnic groups (Table 1),
being lowest in adult Chinese and Japanese, who retain just 5 – 10%
they had as a suckling infant. Whereas in Europeans that lose their
lactase after weaning, the eventual level maybe 30 – 50% they had
as a baby. However it is not entirely clear what these numbers
mean. Human lactase activities are measured from biopsies, and
are expressed as Uymg intestinal weight or ymg protein, or as a
ratio against sucrase-isomaltase. Let us suppose that these values
are only 5% of values found in a suckling infant, i.e.1y20. What
matters is the total activity in the entire small intestine. The length
and surface area, and thus the number of lactase expressing cells, of
the small intestine of an adult will be much greater than when they
were just a few days old. Since the size of the entire small intestine
of an adult is likely to be some 20 times that of the suckling infant,
then if the activity per unit weight is 1y20 that as a suckling infant
then the total level of lactase throughout the small intestine will be
the same. A suckling infant ingests perhaps 1 litre of milk a day,
more than even most white Northern Europeans. This simple
calculation shows that if an adult has lost 95% of their total
lactase, the level per unit weight or protein would have to be just
1% of that in a suckling infant. This raises the question as to
whether the concept of hypolactasia is flawed. These variations
confuse the definition of phenotype in genetic studies, particularly
if symptoms after a lactose load, rather than enzymatic activity, are
used as the principle criterion. The small intestine is made up of
three main segments. The duodenum connects from the stomach.
This then leads to the jejunum, and then the longest section, the
ileum, which then connects to the large intestine. Biopsies are often
taken from the jejunum. But does someone who is highly sensitive
to lactose have a major loss in the ileum? We need a PET or MRI
indicator that can assess the total lactase in the whole of the small
There have been extensive attempts to discover a polymorphism,
and thus a molecular mechanism, to explain loss of lactase on
weaning. It is assumed that lactase persistenceynon-persistence is a
polymorphic trait, where the allele frequencies have been affected
by selection, but where genetic drift has also occurred to influence
haplotype frequency in any particular population. Haplotype is a
single genetic unit on one chromosome, i.e. one member of a pair of
alleles. The unimodal distribution of lactase levels in infants
174 Anthony K. Campbell, Jonathan P. Waud and Stephanie B. Matthews
Black plate (175,1)
moving to a trimodal distribution in adults, with the occurrence of
lactose intolerance in monozygotic twins, support the case that
lactase persistence is a dominant inherited trait, with the genes on
both chromosomes expressing. Lactase persistence is most common
in North West Europe, with highest levels being in the Swedes and
Danes. The mean population level of lactase decreases moving
south. A similar southerly decline is seen when comparing North
and South India. Several ethnic groups and races, known to be
lactase non-persistent, still have some milk in their diet. These
include the Mongols and several groups in Africa the Herero,
Nuer and Dinka tribes. Cows are sometimes retained as a status
symbol; e.g. the Dinkas and Hindus. But the ‘milk’ is often ingested
as a fermented product such as yoghurt or cheese where lactose
levels are much lower than in milk.
Several polymorphisms have been found in the introns and exons
of the lactase gene and its promoter, but none consistently correlate
with lactase persistenceynon-persistence
. There are four common
haplotypes world wide, designated A, B, C and U. Only A, B and C
are found in Europe, A being found in 480% northern Europeans.
The four haplotypes, A, B, C, and U are not related and have
different distributions. The A haplotype has high frequencies only
in the Northern European population, which has a high prevalence
of lactase persistence. The U haplotype is virtually absent in the
Indo-European population. The haplotypes appear to be in a large
region of linkage disequilibrium, where there is evidence of genetic
drift in evolution, but they do not help in identifying the true basis
of lactase persistence. Both alleles from each chromosome express
high levels of mRNA in homozygous lactase persistence. Those
who are homozygous for lactase non-persistence express low levels
from both chromosomes. Heterozygotes express high levels from
the chromosome with the lactase persistent allele, and low levels of
mRNA from the other chromosome. The key question therefore is:
what is the cellular basis of this? Does each cell only express lactase
from one of the chromosomes?
An apparent breakthough was reported by a Finnish group
Two polymorphisms were found in introns of the helicase MCM6,
14 Kb upstream (on the reverse DNA strand, like lactase) from the
lactase gene itself, CyT
in intron 13 and GyA
in intron
9, numbered from the ATG start codon of lactase gene. CC and
GG homozygotes had the lowest level of lactase. Homozygote
TTyAA had full levels of lactase, with heterozygotes being in the
middle. Encouragingly, there have now been several clinical
studies, including our own
, showing that these polymorphisms The molecular basis of lactose intolerance 175
Black plate (176,1)
provide a useful addition to clinical management. There are there-
fore five possible genotypes: CCyGG, CCyGA, CTyGA, CTyAA,
and TTyAA. In our initial analysis, 210 patients referred with
unexplained gut and other problems were investigated. 14.5% were
homozygous CCyGG, 39% were heterozygous CTyGA and 46.5%
were homozygous TTyAA. One patient only was CCyGA, and
responded as the CCyGG. All CCyGG were diagnosed as lactose
intolerant, 83% of CTyGA and 73% of TTyAA. In the control
group, with no history of gut or systemic symptoms, none were
CCyGG, 13% were CTyGA and 87% TTyAA. Although there
have been reports that these polymorphisms can regulate lactase
expression in vitro
, these data do not support the hypothesis that
either of the two polymorphisms are mechanistically the cause of
hypolactasia. Several lactose intolerant families were TTyAA, and
both Finnish and Italian studies had individuals who were CC and
lactase persistent, or TT who were lactase non-persistent. Also
there appears to be no correlation between the expression of
mRNA for MCM6 and lactase in the gut cells of individuals with
hypolactasia or lactase persistence. There are two explanations for
1. The CyT and GyA polymorphisms are simply a closely linked
marker to lactase persistenceynon-persistence.
2. There is genetic heterogeneity causing lactase persistenceynon-
persistence: i.e. there is more than one mutation that causes
lactase persistenceynon-persistence.
Lactase is synthesised in specific cells that begin their life by
division from stem cells in the cleft of the villus in the small
intestine. The intestine is made up of rows and rows of finger like
projections. These are called villi, and are small folds along the
intestine, which are lined by cells. As the cells move up each villus,
the lactase gene is switched on and the lactase product is processed
so that it appears on the apical surface. The cells are scattered in a
non-clonal manner. They exhibit the ‘Rubicon’ principle
,i.e. the
ultimate level of lactase in the entire small intestine depends on the
number of cells expressing lactase, rather than the level of lactase
itself in each cell. If you have only 10% of the lactase you had as a
suckling infant you are likely to have only a small % of the cells
expressing fully lactase that you had as a baby. The final level of
lactase in the intestine would be further reduced by down-regula-
tion regulation of transcription. This implies a fascinating devel-
opmental mechanism, perhaps involving DNA methylation, since
thousands of gut villi cells expressing lactase are replaced every
176 Anthony K. Campbell, Jonathan P. Waud and Stephanie B. Matthews
Black plate (177,1)
day. As any mother will tell you, there is a very simple biological
process that takes a baby off the breast – the appearance of teeth.
Thus, an obvious candidate for switching lactase-cells off would be
deciduous dental homeobox genes, such as bmp-4, msx-1 and -2,
shh, dlx-1 and -2, and lef1
. We have also observed that children
who have a parent diagnosed as lactose intolerant, seem to become
sensitive to lactose as they get their secondary teeth, and can
become fully lactose-sensitive after puberty.
How failure to digest lactose leads to symptoms
Symptoms occur when lactose, undigested in the small intestine,
reaches bacteria in the large intestine. These bacteria metabolise
lactose, producing gases that distend the gut, causing pain and
flatus, and toxins that, when reabsorbed into the body, cause
harmful effects on a range of tissues, including neurones, heart
cells, other muscles, endocrine cells, and cells of the immune
There are two ways in which lactose can be prevented from being
digested in the small intestine:
1. Insufficient lactase.
2. Insufficient monosaccharide uptake after lactose cleavage.
Insufficiency arises either because there is not enough protein, as in
hypolactasia, or from inhibition by something in food. Lactase is
inhibited competitively and non-competitively by a number of
naturally occurring substances. But, none have yet been shown to
be involved in lactose intolerance. However, the uptake of galac-
tose and glucose, through the Na
dependent transporter SGLT1
at the apical surface of the enterocytes in small intestine, can be
inhibited by several substances found in food. SGLT1 enables
monosaccharides to be transported into cells against a concentra-
tion gradient, using the Na
gradient as an energy source. The
glucose and galactose are then transported into the blood at the
other side (basolateral) of the cell by another glucose transporter
not dependent on sodium called GLUT2. Galactose is quickly
metabolised by the liver, as it is toxic to the eye, and other cells.
This pathway is inhibited by ethanol, hence the use of ethanol in
early lactose tolerance tests using measurement of blood galactose
andyor glucose as an indicator. SGLT1 is inhibited by the tri- and
tetra-saccharides, raffinose and stachyose
. These are found in
beans, pulses, root vegetables such as parsnips, and chick peas used
in the production of humus. These sugars cause gas and toxins in The molecular basis of lactose intolerance 177
Black plate (178,1)
the large intestine because not only are they not broken down in the
small intestine, but also they inhibit the uptake of glucose and
galactose. So glucose from starch or lactose hydrolysis ends up in
the large intestine
Not all sugars are transported into the gut epithelial cells by
SGLT1. Fructose uses another transporter, facilitated diffusion via
GLUT5 that does not use the Na
gradient. Intracellular fructose,
like glucose, is then transported into the blood using GLUT2 on
the other side of the cell. GLUT5 can be overloaded, and may be
inhibited by certain substances found in food and drinks. Fructose
tastes sweeter than glucose or sucrose. Hence, evolution has
produced it as the main sweetener in fruits such as apples and
grapes. One of our patients became ill after drinking two glasses of
home made apple juice. We estimated that she had drunk the
equivalent of 20 apples! As with lactose, a fructose industry has
grown up over the past few decades, adding it as corn-syrup to
sweeten many foods and drinks. Could other compounds, natural
or food additives, also inhibit SGLT1 or GLUT5. One candidate is
bcoumarin (Figure 2), the orange colour in orange juice, since
several patients complain of a headache 2 – 3 h after drinking
orange juice. Grapefruit contain a substance that interacts with
The bacterial toxin hypothesis
Lactose itself, and galactose, could be toxic if absorbed into the
blood stream. But the major cause of symptoms in food intolerance
is the production of gases and toxins by gut bacteria. The large
intestine contains some 10
individual bacteria, 100 times the cells
in the rest of our body. There are over 1000 different species. The
level of oxygen in the large intestine is low, probably 51mM, 1y200
of that in air-saturated water. Thus, 490% of the bacteria there
are anaerobes. At least 25% are Bifidobacter, with the rest being
other strict anaerobes. Some, such as Bacteroides, are so sensitive
to oxygen that they are very difficult to culture directly from gut
samples, as they die immediately on exposure to the air. The
remaining eubacteria are mainly facultative anaerobes. Less than
1y1000 of gut bacteria are aerobes. There are also archaebacteria,
responsible for methane production. And there can be yeasts such
as Candida and fungi.
Bacteria release a wide range of substances (Table 2). In low
oxygen, bacterial metabolism of lactose and other carbohydrates
produces gases and a range of small organic molecules. It is
178 Anthony K. Campbell, Jonathan P. Waud and Stephanie B. Matthews
Black plate (179,1)
absorption of these that cause the symptoms of lactose intolerance.
In order to make ATP, anaerobic bacteria use substrate level
phosphorylation instead of oxidative phosphorylation. If the
NADH from this is not re-oxidised to NAD, then glycolysis will
shut down. In exercising muscle, we do this by generating L-lactate:
HþþNADH þpyruvate ?NADþþLlactate ð3Þ
Anaerobic bacteria have evolved several other ingenious pathways
in order to remove the H from NADH, through ‘fermentation’
(Figure 5). Several gut bacteria generate D-lactate instead of L-
lactate. Measurement of blood D-lactate can thus be a good
indicator of bacterial activity, e.g. in stressed neonates. A major
route for removing the H from NADH is through the generation of
gases, the cause of flatus. The main gas is H
, with some CH
the archaebacteria. Many bacteria contain an inducible formate
hydrogenase, discovered by Marjory Stephenson in the 1930’s,
converting formate into CO
and H
The H
can then act as substrate for the methanogenic archae-
This explains why in some patients CH
is a useful clinical
indicator, when lactose ingestion results in little or no H
in the
breath. In this case, H
has been converted to methane. H
and The molecular basis of lactose intolerance 179
Table 2 Substances that can released by different bacteria and archaebacteria
Product released Example
Gases Carbon dioxide, hydrogen, methane, hydrogen
sulphide, oxygen, nitrogen, ammonia
Ions Calcium, sodium, potassium, magnesium,
manganese, iron
Metabolites Alcohols, diols, aldehydes, short chain fatty acids,
dimethyl hydrazine, amino acid degradation
products, cyclic AMP
Vitamins K, B12, thiamine, riboflavin
Pheremones Lactones, cytokines
Small molecule toxins Antibiotics, tetrodotoxin
Drugs Many
Peptides Toxins, enzymes
Nucleic acids Competence factors, plasmids, bacteriophages
Polymers Poly hydroxybutyrate
Any particular bacterium can only release some of these.
Black plate (180,1)
180 Anthony K. Campbell, Jonathan P. Waud and Stephanie B. Matthews
Fig. 5. The pathways generating gases and putative toxins by anaerobic bacteria in the large intestine. The structures of the fermentation products
and putative toxins are shown in the box.
Black plate (181,1)
are therefore the main gases in flatus, with some H
S from
sulphurous bacteria. Absorption of these gases into the blood
allows them to be detected in the breath.
In addition, there are several other pathways for removing the H
from NADH, generating alcohols, diols, aldehydes, ketones and
acids. These include acetaldehyde, acetoin, butan 2, 3 diol,
dimethyl glyoxal, diacetyl, ethanol, formate, methane, propan 1,
3 diol and short chain fatty acids (Figure 5). Several of these have
been detected in blood samples taken during a lactose tolerance
test, indicating that colonic bacteria are actively metabolising
lactose. Ironically, the one enzyme you don’t want to see its
substrate is the b-galactosidase in bacteria, whose induction by
lactose lead to the discovery of mRNA by Jacob and Monod,
heralding the DNA revolution. Because once the b-galactosidase in
the bacteria of the large intestine sees lactose, it metabolises it to
gases and toxins.
A crucial group of putative toxins are the diols. Butane 2, 3 diol
is a fermentation product of glucose, there being three naturally The molecular basis of lactose intolerance 181
Fig. 6. The effect of the bacterial fermentation products butane 2, 3 diol and
propan 1, 3 diol on cytosolic free Ca
in E. col. JM109 E. coli cells expressing
the bioluminescent Ca
indicator aequorin were incubated in 25 mM HEPES,
125 mM NaCl, 1 mM MgCl
, pH 7.5 for 1 min in a luminometer, and the
luminescence counts recorded
was then added for 2 min and then
100 mM meso butane 2, 3 diol or propane 1, 3 diol added for a further 5 min.
Cytosolic free Ca
was then estimated by converting the luminescent counts to
free Ca
using a standard curve. Results represent the mean ¼ySEM from
eight separate experiments. Temperature 21C. (j)¼100 mM butane 2, 3 diol;
(m)¼100 mM propane 1, 3 diol; (d)¼Control with no diol.
Black plate (182,1)
occurring stereoisomers: meso, 2R, 3R ( ), 2S, 3S ( þ). Propan 1,
3 diol is a fermentation product of glycerol. Harden and Walpole
showed that fermentation products of Aerobacterer aerogenes
differed from those produced by E.coli, consisting mainly of
butane 2, 3 diol and acetoin. The production of acetoin, and its
oxidation product diacetyl, is the basis of the Voges-Proskauser test
widely used in bacteriology. Other bacteria capable of producing
butane 2, 3 diol include: Klebsiella,Enterobacter,Serratia,Bacillus,
Lactobacillus, and Aeromonas, all of which can be found in the
human colon, though butane 2, 3 diol is not the only fermentation
product. Three enzymes are required to produce butane 2, 3 diol
from pyruvate: a-acetolactate synthase, a-acetolactate decarboxy-
lase and acetion reductase. In Enterobacter and Klebsiella, butane
2, 3 diol production in culture requires an acid pH and the presence
of acetate as a regulator.
The plasma concentration of butane 2, 3 diol in healthy humans
or alcoholics is 10 – 100 mM. If the lactose in a glass of milk
(approx. 10 g) were converted to butane 2, 3 diol, then the local
concentration of this diol in the gut would be 100 – 200 mM. Butane
2, 3 diol, and propane 1, 3 diol, generate Ca
transients in E. coli
(Figure 6). The role of cytosolic free Ca
as an intracellular signal
is well established in eukaryotic cells
. However, the role of
intracellular Ca
in bacteria is less well established
transients and effects of diols on growth (Figure 7), suggest that
sugar fermentation products may determine the balance of
bacterial species in the colon. Changes in gene expression that
lead to just a 10% decrease in generation time would, through
Darwinian-Wallace selection, result in 490% of these bacteria
dominating within 20 generation times, i.e.524 h.
Other bacterial toxins include amino acid degradation products
such as the phenol cresol, indoles and skatoles (Figure 2), or
peptide and protein toxins. The bacterial toxins are primitive
signalling molecules. They act on pathways that switch cells on
or off in the nervous system, heart and muscles, and the immune
system. Butane 2, 3 diol also appears to affect Ca
signalling in
the cytosol and ER of tissue culture cells, and apoptosis (Trimby
and Campbell, unpublished). In a model invertebrate system – the
water flea Daphnia – lactose induces heart arrhythmia
, similar to
that in 25% of our patients with lactose intolerance.
The idea that bacteria in the gut can release toxins is over 100
years old. Elie Metchnikoff (Figure 8) was a founder of modern
immunology, discovering phagocytes for which he was awarded
one of the earliest the Nobel Prizes with Ehrlich in 1902. However,
182 Anthony K. Campbell, Jonathan P. Waud and Stephanie B. Matthews
Black plate (183,1)
his real intellectual ‘baby’ was the idea that gut bacteria produce
toxins. He wrote; ‘The large intestine must be regarded as one of
the organs possessed by man and yet harmful to his health and his
life. The large intestine is the reservoir of the waste of the digestive
processes, and this waste stagnates long enough to putrefy. The
products of putrifaction are harmful.’ ‘Bacterial putrefaction is the
cause of all disease.’ He even carried out experiments injecting
cresol and other putative toxins into mice, showing they could be
A key issue is whether sugars such as lactose can induce gene
expression and growth of toxin-producing bacteria, as opposed to
those that simply produce gas. And also whether there are just one
or two species of bacteria capable of producing large amounts of
toxins, analogous to Helicobacter whose discovery revolutionised
the treatment of stomach ulcers. The molecular basis of lactose intolerance 183
Fig. 7. The effect of butane 2, 3 diol on bacterial cell growth. JM109 cells
expressing aequorin were suspended in 25 mM HEPES, 125 mM NaCl, 1 mM
, pH 7.5. 50 ml aliquots were added to 15 ml of growth medium LB (Luria
Bertani) medium (1% tryptone, 0.5% yeast extract and 0.5% NaCl pH 7.2)
with carbenicillin (100
) and 5 mM EGTA with or without butane 2, 3
diol or 10 mM Ca with or without butane 2, 3 diol. The cells were then incubated
at 37C with vigorous shaking for up to 24 h and samples taken every hour to
measure viability by their ability to grow as assessed by the absorbance at
600 nm. The generation times were then estimated. Colony counts confirmed the
viability of the cells under all conditions. Results represent the mean þySEM
of three determinations. Statistical significance: 5 mM EGTA þ100 mM butane
2, 3 diol versus 10 mM Ca
þ100 mM butane 2, 3 diol P¼0:0002;5mM
EGTA versus 5 mM EGTA þ100 mM butane 2, 3 diol P¼0:008;10mMCa
versus 10 mM Ca
þ100 mM butane 2, 3 diol P¼0:09. All other comparisons
Black plate (184,1)
The science of clinically managing lactose
Irritable bowel syndrome (IBS), with unexplained gut problems –
pain, distension, gas, tummy rumbling, diarrhoea or constipa-
tion – is the most common problem faced by gastroenterologists.
But patients may also complain of non-gut (systemic) symptoms,
including severe recurrent headaches, chronic fatigue, loss of
concentration and a dizzy head, muscle and joint pain, allergies
such as eczema, pruritis, urticaria, asthma, sinusitis, rhinitis and
184 Anthony K. Campbell, Jonathan P. Waud and Stephanie B. Matthews
Fig. 8. Elie Metchnikoff (1845 – 1916), pioneer of the bacterial toxin
Black plate (185,1)
hay fever, heart palpitations, and increased micturition
1 – 3,45
(Table 3). It is these non-gut symptoms, and their irregular
occurrence, that have confused diagnosis. A high percentage of
these patients are intolerant to lactose
. In some cases this explains
all their symptoms, while others have a wider food intolerance
often to other carbohydrates such as fructose and starch in
particular forms, and foods containing stachyose or raffinose
(Figure 1). Their threshold to these non-lactose foods varies
considerably, confusing diagnosis and treatment.
We have now analysed data from several hundred patients
referred to our food intolerance clinic, the first in Wales. Our
recommended diagnosis and management of lactose intolerance
now is
: The molecular basis of lactose intolerance 185
Table 3 Gut and systemic symptoms of people with lactose intolerance
Symptoms of lactose
No. of people with
(% of total with lactose
A. Gut related
Abdominal pain 100%
Gut distension 100%
Borborygmi (tummy rumbling) 100%
Flatulence (gas) 100%
Diarrhoea 70%
Constipation 30%
Nausea 78%
Vomiting 78%
B. Systemic
Headache and light headedness 86%
Loss of concentration and poor short term
Chronic severe tiredness 63%
Muscle pain 71%
Joint pain, andyor swelling and stiffness 71%
Allergies, such as: 40%
Eczema (skin rash)
Pruritis (itchy skin)
Rhinitis (runny nose)
Sinusitis (stuffed up sinus)
Asthma (wheezing and shortness of breath)
Heart arrhythmia 24%
Mouth ulcers 30%
Increased frequency of micturition (weeing) Less than 20%
Sore throat Less than 20%
Systemic ¼around the body.
Black plate (186,1)
1. Buccal (swab inside the mouth) sample for CyT
analysis (Figure 9).
a. If CC, immediately remove of all lactose from diet. If
symptoms improve after one month diagnosis of lactose
intolerance confirmed.
b. If CT or TT carry out lactose tolerance test.
2. New recommended lactose tolerance test:
a. 50 g (1 gykg for children) dissolved lactose.
186 Anthony K. Campbell, Jonathan P. Waud and Stephanie B. Matthews
Fig. 9. Polymorphism analysis in lactose intolerance. PCR was carried out on DNA isolated from buccal (mouthy
swab) samples as described in ref. 2. Introduction of restriction sites enabled samples homozygous for CCyGG,
heterozygous for CTyGA or homozygous for TTyAA to be distinguished.
Black plate (187,1)
b. Record breath hydrogen and methane for 6 h.
c. Record all symptoms for 48 h.
d. If the breath test is positive, i.e.H
rises to 420 ppm or CH
45 ppm over the nadir, then change to a lactose free diet.
e. Every patient followed up in 12 weeks for a definitive
f. If the breath test is negative, but there is a significant increase
in symptoms after the lactose load, the patient should
undergo a supervised trial to determine their lactose thresh-
g. Family studies should be carried out to determine other
affected individuals.
h. Hypolactasia caused by infections such as Giardia or rota-
virus should be investigated if there is no evidence of family
3. Give advice on lactose free meals, and the danger of hidden
4. Follow up in 1 year.
5. Calcium and vitamin D status should be monitored, and advise
on the use of probiotics.
6. Patient advised to keep a food diary to identify culprits if caught
Several hundred patients with unexplained gut and other symptoms
have now been referred by GPs and consultants to our clinic. The
patients were diagnosed using the new clinical procedure into those
with lactose intolerance and those not lactose intolerant. Those
diagnosed without lactose intolerant were all CyTorTyA. But
there was no difference in the total number of symptoms reported
using CyT
genotyping between the lactose intolerant and non-
lactose intolerant groups. However, a major difference was found
between these two groups when lactose was removed from the diet
(Figure 10). 100% of CCyGG, 83.3% of CTyGA patients and
76.3% TTyAA were diagnosed as lactose intolerant. Thus for
CCyGG a breath test is unnecessary. This is of considerable
benefit, as many suffer badly from prolonged symptoms after the
50 g lactose load used in the lactose intolerance test. This major
revision of the clinical management of lactose intolerance has not
only benefited individual patients, but has resulted in huge savings
for the NHS. Many of our patients were constantly seeing their
GPs and specialists, and were taking a cohort of drugs. Most are
now off all drug therapy and rarely have to see a doctor! Coming
off lactose reduced the number of symptoms from an average of
nine to one (Figure 10). The molecular basis of lactose intolerance 187
Black plate (188,1)
Probiotics are friendly bacteria such as Lactobacillus that can be
taken with food and are claimed to help digest foods when someone
has an intolerance. But the scientific evidence for long term benefits
as opposed to short term placebo effects is weak. There is a
suggestion that dietary intake of lactose prior to the test may
have an inverse affect on the breath test
. This is consistent also
with the increased sensitivity experienced by some patients when
they eliminate lactose from their diets, while higher lactose intake
prior to the breath test may reduce symptoms and gas score, when
compared to patients who have a low lactose intake prior to the test
(50 g lactose). This again highlights the limitations of the current
breath test.
In our cohort of 4250 patients, all referred with unexplained gut
and other symptoms, the percentage of total symptoms (abdomi-
nal þsystemic) reported during the lactose tolerance test was
significantly higher in lactose intolerant individuals than those
who had gut symptoms but turned out to be lactose tolerant
(440%, P¼0:01, n¼130). However no significant difference
was observed when total symptoms were compared by CyT
lactase genotyping alone (P¼0:1). In contrast significant differ-
188 Anthony K. Campbell, Jonathan P. Waud and Stephanie B. Matthews
Fig. 10. Number of symptoms before and after a lactose tolerance test.
Individuals were given 50 g lactose (1 gykg for children), and symptoms recorded
with severity, using a standard clinical scale from 0 – 10, for 48 hours. Breath
hydrogen and methane were also measured every 30 min for 6 hours. The number
of symptoms was recorded prior to the test, during the test and after 1 month on a
lactose free diet. These were plotted, after separating the patients into the
polymorphism groups CCyGG, CTyGA and TTyAA. Control subjects were
normal volunteers with no history of gut and other symptoms.
Black plate (189,1)
ences were observed between lactase genotype when symptoms
were categorised to abdominal, neuromuscular, cardiac, oral and
allergy in lactose intolerant and tolerant patient groups (P¼0:01).
Palpitations (cardiac) symptoms were only observed in lactose
intolerant patients, being highest in CT genotypes, a finding
consistent with our observations that hyperlipidaemia is more
prevalent in lactose intolerant CT genotypes. Oral symptoms
such as mouth ulcers were significantly higher in intolerant patients
when compared to tolerant, being highest in the CC genotype, with
a low oral symptom prevalence in tolerant TT genotypes and no
reports from CT genotype tolerant individuals. This is consistent
with our hypothesis that intestinal bacterial toxins and high levels
of hydrogen and methane gas lead to epithelial hypersensitivity and
ulceration. Anal hypersensitivity is well described in irritable bowel
syndrome but with unknown cause. Constipation was only
reported in the lactose intolerant patient group. General allergy
symptoms were also markedly higher in the intolerant patients
when compared to tolerant. Abdominal symptoms were higher
(435%) in lactose intolerant patients compared to those with
symptoms, but who turned out to be lactose tolerant, but not
distinguishable by genotype. Similar findings were observed for
neuromuscular symptoms, including muscle and joint pain (465%
higher in lactose intolerant).
The differences in type of symptom in referred patients even-
tually diagnosed as lactose intolerant emphasise the benefit of
genotyping in the clinical management of this condition.
The problem of lactose in food
Dairy products, together with foods and drinks containing milk,
are abundant in the supermarket. Some are obvious, others are not.
Food labelling is poor. Many patients do not realise that if dried
milk powder, condensed or evaporated milk is used, then this will
add more lactose than the equivalent amount of milk (Table 4).
Recipe books sold with home bread makers recommend adding
dried milk powder to many recipes. Many Asian restaurants are
now using cream and evaporated milk instead of the classic
ingredient in Asian cooking, coconut milk which is lactose free.
Then there is the problem of pharmaceuticals. The major filler is
usually lactose. And many fluids used clinically in enteral feeding
contain lactose. Many people do not realise that products such as
whey contain all the lactose in milk. In order to isolate casein, the
major milk protein, the milk is first centrifuged and the cream The molecular basis of lactose intolerance 189
Black plate (190,1)
190 Anthony K. Campbell, Jonathan P. Waud and Stephanie B. Matthews
Table 4 Lactose content of some foods and drinks
Food or drink Relative lactose content Lactose as a %
(i.e.gy100 g or 100 ml)
Lactose Complete 100%
Whey Very high 70%
Non-fat dry milk powder Very high 50%
Cow’s milk High (47 gylitre) 5%
Goat’s milk High (44 gylitre) 4%
Reduced lactose milk Low 1%
Lactose-free milk Very low 0 – 0.5%
Sour milk High 4%
Buttermilk High 4%
Commercial yogurt High 4%
True natural yogurt Moderate to low 2%
Cheese Moderate
Feta Quite high 4%
Diet cottage Quite high 3%
Parmesan (hard block for fresh grating) Low to moderate 1%
Parmesan (grated in packet) Can be quite high 3%
Cheddar Trace to low 0– 2%
Camembert Not very high 0 – 1%
Edam Low 0– 1%
Some cheese products Can be high 5 – 10%
Cream Moderate 4
Butter Low 1
Clarified butter Very low Should be 0
Chocolate High ?
Milk proteins (casein etc.) Low, but can be present 0 – 2
Black plate (191,1) The molecular basis of lactose intolerance 191
Table 4 Lactose content of some foods and drinks (continued)
Food or drink Relative lactose content Lactose as a %
(i.e.gy100 g or 100 ml)
Non-dairy ‘hidden’
Processed meats such as sausages and salamis Added, can be high ?
Breads and cake mixes Added, can be high ?
Slimming bars Added, can be high ?
Powdered sauces Added, can be high ?
Reduced fat foods e.g. mayonnaise and biscuits Added, can be high ?
Lager Added, can be high ?
Powdered or artificial fruit juice Added, can be high ?
Fresh meat from the butcher None 0
Fresh fruit None 0
Fresh vegetables None 0
Eggs None 0
Pure squeezed orange juice None 0
Lactic acid (lactate) None 0
The numbers in this table have been rounded up and thus are very approximate. ? ¼no accurate levels known. 1% is equivalent to about 20 ml
milk, a typical amount in a cup of tea, 10% to 200 ml. A normal glass of milk contains about 200– 250ml, a block of butter weighs 250 g, and
a block of cheese perhaps 250 500 g. A spoonful of Parmesan cheese, freshly grated ¼about 30 g, equivalent to 0.3 g lactose or 6 ml milk.
Black plate (192,1)
skimmed off the top. The remaining fluid, the skimmed milk, is
acidified to a pH of 4.7 so that the casein precipitates. The
precipitate is removed leaving the supernatant whey. This
contains 20% of the original protein in the milk, and crucially all
the lactose. Whey is added to many foods, but lactose itself may not
be on the label. It is the whey that is used to make lactose itself, by
evaporation to crystallisation.
A further major problem is that of ‘hidden’ lactose (Table 4).
Lactose can be added to breads, cake mixes, sausages and
processed meats, and even chicken and drinks, without being
properly labelled
. Labelling regulations have changed in Europe
since 2005. But many food manufacturers have been slow to
respond to this. The US alone produces some 300 million kg of
lactose per year (Figure 11). Everyone can tolerate some lactose.
You would have to eat a kilogram of Parmesan cheese to be
equivalent to a glass of milk. Sprinkling a teaspoon on pasta
should therefore be no problem. But, the amount of ‘hidden
lactose’ can be considerably more than this. We estimated that
the amount of lactose in a slimming drink taken daily by one of our
patients was equivalent to 1 – 2 litres of milk!
Darwin’s illness revealed
‘I have had a bad spell, vomiting every day for eleven days and
some days after every meal’. So Charles Darwin (1809 1882) wrote
in a letter to his friend Joseph Hooker in December 1863. Later he
wrote to his father, a doctor himself, ‘The sickness starts usually
two hours after a meal’. In fact Darwin had already suffered chest
pain and heart palpitations in December 1831 while staying in digs
192 Anthony K. Campbell, Jonathan P. Waud and Stephanie B. Matthews
Fig. 11. The production of lactose in the USA over the past 15 years and
Black plate (193,1)
at Plymouth awaiting better weather for the Beagle to depart. He
told no one until years afterwards for fear he would not be allowed
on his ‘trip of a lifetime’. For over 40 years Charles Darwin was
frequently ill. He lived in the Kent village of Down(e) as a semi
recluse because he was ill so much, sometimes for days on end. He
failed to go to the famous Oxford debate in 1860 because he was in
the middle of one of his attacks. He saw some twenty doctors,
including his father, and tried dozens of remedies None really
worked, though Darwin did seem to improve when he underwent
Gully’s water therapy at Malvern. The only time he got better was
when, by chance, he came off milk.
Darwin’s symptoms fit exactly systemic lactose intolerance
(Table 5). Darwin suffered from stomach ache, flatulence, head-
aches and a swimming head, vomiting, and chronic fatigue, joint
pains, skin rashes and boils, mouth ulcers and heart palpitations.
And he was often depressed. Many proposals have been put
forward to explain his illness, including arsenic poisoning,
Chagas’ disease and psychosomatic disorders such as bereavement
syndrome, because of the death of his mother at the age of 8. None
match his symptoms. Six pieces of evidence support our hypothesis
that Charles Darwin suffered from lactose intolerance:
1. Darwin’s symptoms fit exactly those we have identified in
systemic lactose intolerance.
2. The timing of his vomiting and gut pain was 2 – 3 hours after a
meal, just as expected for lactose to reach the large intestine.
3. His wife Emma used milk and cream constantly in her recipes.
4. There was a clear history of illness in the Darwin family, in his
children and on the Wedgwood side of the family.
5. Darwin did not suffer from his illness on the Beagle (1831
1836) where there was no fresh milk. He just had sea sickness
and a fever in South America, probably typhoid.
6. Darwin only got better when, by chance, he came off milk.
What Darwin missed
Charles Darwin not only missed the cause of his life-time illness,
but he also missed the most important aspect of our own evolu-
. In ‘The Origin’ one chapter (4 in the 1st edition of 1859, and
6 in the 6th edition of 1868) is entitled ‘Difficulties onyof Theory’.
These were not the famous difficulties highlighted by his oppo-
nents. Rather they were Darwin’s difficulties. He wrote, Natura non
facit saltus(m) – Nature takes no leaps. He saw no way to explain The molecular basis of lactose intolerance 193
Black plate (194,1)
194 Anthony K. Campbell, Jonathan P. Waud and Stephanie B. Matthews
Table 5 Systemic lactose intolerance versus Darwin’s illness
Symptoms of systemic
lactose intolerance
% people with
lactose intolerance
who have this
Darwin’s description
of his symptoms
of Darwin’s
Gut symptoms (pain, bloating,
100% Stomach ache Common
Flatulence (farting) 100% Flatulence (belching) Common
Headache 86% Headache Common
Light headedness and loss of
82% Swimming head and difficulty to
Nausea and vomiting 78% Vomiting Very common
Muscle and joint pain 71% Rheumatic pain Often
Tiredness and chronic fatigue 63% Chronic fatigue and exhaustion Very common
Allergy (eczema, hay fever,
rhinitis, sinusitis)
40% Skin rash and boils Often
Mouth ulcers 30% Mouth sores Common
Heart palpitations 24% Palpitations in the chest Common
Depression Common,
but not quantified
Depression Frequent
*Proportion of people diagnosed as lactose intolerant who have this particular symptom within 48 h of taking lactose. Darwin’s occurrence
is based on his notes and letters during periods of the attacks.
Black plate (195,1)
how small change by small change could lead to the origin of the
electric organs of fishes, the luminous glands of fireflies and glow-
worms, and even the eye. Yet he failed to highlight the most
obvious Rubicon crossed by our evolutionary ancestors – the
breast and its ability to produce milk. Even in ‘The Descent of
Man’ this aspect of human biology has just a cursory mention. So
does the principle of natural selection alone explain this unique
feature in the evolution of mammals, and our own species? Which
came first, lactose or lactase? How does a new protein such as
lactase, or a process such as lactose production, originate and
develop before it can respond to the forces of natural selection?
Lactose is restricted to the milk of terrestrial mammals, but
cerebrosides (glycosyl ceramides) are present in the diet of all
vertebrates. The origin of intestinal lactase is therefore likely to
be its phlorizin, or rather its glycosyl ceramidase, activity. Of the
common sugars found in plants, the order of sweetness is fructo-
se4glucose ¼sucrose4lactose (b4a). Milk is not sweet because
lactose has 1y6 sweetness of sucrose. A non-sweet sugar would be
much less prone to attracting insects to the breast. Then molecular
biodiversity took over
the evolution of the diversity of lactase
levels within the human population.
Domestication of animals and agriculture, and cheese-making,
began some 10,000 years ago
. Legend has it that an Arabian
merchant was carrying a pouch made of sheep stomach, full of
milk. The heat of the sun, together with the release of rennet from
the stomach, caused the milk to separate into the solid curds and
the liquid whey. Rennet contains the protease rennin (not to be
confused with renin). This cleaves a glycopeptide from casein to
form paracasein, which then binds Ca
, causing the protein to
precipitate to form the curd. Dairying proper did not begin until
6,000 – 8,000 years ago, originating in the great civilisations of
Babylon and Assyria from Mesopotamia
. The use of milk
probably began with camels and goats. This was followed some
1,000 – 2,000 years later by the use of milk from sheep and cows.
Archaeological data from pots and other artefacts, together with
ancient writings, seals and drawings, puts the origin of milk
drinking even more recent, in Mesopotamia and Egypt 5,000
years ago, and in Africa 7,500 years ago.
10,000 years ago a huge geological change occurred that had a
major influence on the current prevalence of lactase persistence in
the white Northern Europeans. The last ice age ended, having
begun in the Pleistocene Era 2.5 million years before, freeing
Europe from ice. Given a generation time in humans of 20 – 30 The molecular basis of lactose intolerance 195
Black plate (196,1)
years, and an origin of dairying some 6,000 years ago, this means
that there have been only some 200 300 generations to select the
90% prevalence of lactase persistence in Northern Europeans
today. In the nomads of Asia and Africa camel’s milk, cheese
and yoghurt are major components of their diet. Humans moving
north into the plains of Europe would have needed a transportable,
and continuous, food supply.
Three hypotheses have been proposed for the selective advantage
of lactase persistence; i.e. keeping lactase after weaning and thus
being lactose tolerant, rather than lactose intolerant:
1. A major food source for nomadic populations.
2. A source of water in desert zones.
3. A source of calcium in geographical areas where sunlight is
But the real puzzle is, why do all mammals, including most
humans, lose most of their lactase after weaning? Why not keep
it all? Linnus Pauling argued that keeping a protein, such as lactase,
would be energetically wasteful, when there was no dietary source
of its main substrate. But there must be another selective advan-
tage. Non-milk drinking communities have a very different diet
from the first dairying groups of humans. In addition to fish and
meat, the diet of Asians and Africans contains brown and white
rice, soya, beans and pulses, exotic fruits such as bananas, oranges
and lemons, spices and nuts. Most of these only became available
in Europe after the 15th and 16th century voyages of explorers.
Until then, the European diet consisted mainly of dairy products,
animal and bird meat, eggs, a few natural fruits when in season
such as apples and berries, wild herbs, and bread, once agriculture
was in full swing. Many exotic spices and fruits may contain
substances analogous to phlorizin, in that they may be hydrolysed
by lactase to products that are potentially poisonous and patho-
genic. So there would be a clear selective advantage of only keeping
the minimum amount of lactase, necessary to digest the small
amount of glycosyl cerebrosides in the diet.
Attempts to develop mathematical models for lactase persisten-
, producing a population where 480% are
lactase persistent, assume:
1. The ancestral state was lactase non-persistence; i.e. loss of
lactase on weaning.
2. A mutation occurred about 10,000 years ago leading to lactase
196 Anthony K. Campbell, Jonathan P. Waud and Stephanie B. Matthews
Black plate (197,1)
3. Lactase persistence has a Darwin-Wallace selective advantage.
4. This selective advantage is reflected mathematically by a high
selectivity coefficient.
A founder effect with genetic drift seems the most plausible
explanation for the world-wide distribution of the four main
alleles A, B, C and U, and the prevalence of lactase persistence
or non-persistence in particular ethnic or genetic groups, with loss
of certain haplotypes such as U outside Africa. A delay in weaning,
concomitant with retention of lactase, has been proposed to have a
selective advantage in monkeys, as the young would be protected
longer and births of future siblings would be more spaced out.
However this does not explain the selective advantage of retaining
large levels of lactase in white Northern Europeans.
Many hypotheses about the evolution of lactase persistenceynon-
persistence, and mathematical models, are flawed because:
1. Natural selection works on the phenotype in populations, not
the genes of individuals.
2. The mechanism of lactase persistenceynon-persistence involves
first a change in the number of cells expressing lactase
, and
only then regulation of the level of lactase within the cell.
3. There is a huge molecular biodiversity in the level of lactase
within and between genetic and ethnic groups.
4. Natural selection does not take account of the reproduction
5. The numbers of individuals where selection was acting are too
small to allow mathematical models that use probabilities, and
assume populations of ‘infinite’ size, In the Galapagos finches,
evolution of new species occurred through just a few 100
individuals in each generation.
Lactose intolerance illustrates the Rubicon principle
, where a
threshold has to be crossed before biological experience begins.
Five such Rubicons need to be explained in the evolution of lactase,
and lactose intolerance:
.The origin of mammals, lactose and the cells of the mammary
.The origin of lactase and the specialised cells in the small
.The switch that determines whether an intestinal cell expresses
lactase or not.
.The level of lactose that generates significant gas and toxins
when it reaches the bacteria in the large intestine. The molecular basis of lactose intolerance 197
Black plate (198,1)
.The threshold that determines the experience of a particular
Is milk bad for you?
There are reports claiming that, for adults, milk is beneficial, and
may reduce, for example, heart disease
. Others claim that milk
intake correlates with heart attacks, certain types of cancer and
even Parkinson’s disease
. Milk is highly nutritious, containing
proteins, fats, salts, and vitamins. But can any potential harmful
effects of milk be attributed to lactose? The key is mechanism – the
bacterial toxin hypothesis. Only when studies correlating milk
consumption with an end response, such as heart disease or
cancer, identify a mechanism will there be clarity. A further
reason for confusion is the lack of separating cohorts into ethnic
groups, and genetically based on the CyT polymorphism
Without this, any conclusions will be essentially meaningless. A
further problem is that of biochemical individuality
. The
study of lactose intolerance demonstrates the need for a new
approach to epidemiology, where mechanism and individual mole-
cular diversity within a population are taken into account.
The future
Science is about discovering how the Universe works, from the big
bang to how bacteria naturally evolved to be resistant to anti-
biotics. Lactose intolerance highlights a molecular mechanism,
forgotten for 100 years, that is likely to be of major importance
in many unsolved diseases, such as the diabetic epidemic in Asians,
reactive and rheumatoid arthritis, multiple sclerosis and some
cancers. Lactose intolerance also illustrates the need for a new
approach to how we teach medicine, moving away from box ticking
and the heavy reliance on drugs to an understanding of mechanism.
Coming off lactose has transformed our lives and those of our
families, as well as the lives of several hundred patients. They now
feel wonderful, with a massive reduction in drugs and visits to the
doctors, and even coming off surgery lists. Three of our patients
even became unexpectedly pregnant after coming off lactose. It is
hardly surprising Darwin missed his own lactose intolerance. This
condition was not recognised in the 19th century. But how did he
miss this most obvious characteristic in our own evolution? The
science of lactose intolerance can reveal the answer to the problem
198 Anthony K. Campbell, Jonathan P. Waud and Stephanie B. Matthews
Black plate (199,1)
Darwin never really addressed, the true ‘origin’ rather that the
‘development’ of the human species
We thank our colleagues John Green and Jill Swift, Suzanne
Watkins, and our patients. We thank The Department of
Medical Biochemistry and Immunology and the Wellcome Trust
to AKC (Grant no. 075897yZy04yZ), for financial support. We
also thank the Institute of Biomedical Science for a research grant
to JPW.
1. Matthews, S. B. and Campbell, A. K. (2000) When sugar is not so sweet.
Lancet,355, 1330.
2. Matthews, S. B., Waud, J., Roberts, A. and Campbell, A. K. (2004)
Systemic lactose intolerance: a new perspective on an old problem. Post
Grad. Med. J., 81, 167– 173.
3. Campbell, A. K. and Matthews, S. B. (2005) Tony’s Lactose Free Cook-
book – the science of lactose intolerance and how to live without lactose.
(ISBN 0-9540866-1-9).
4. Jacobi, A. (1901) Milk-sugar in infant feeding. Trans. Am. Pediat. Soc., 13,
150 – 160.
5. Kretchmer, N. (1971) Memorial lecture: lactose and lactase – a historical
perspective. Gastroenterology,61, 805– 813.
6. Holzel, A., Schwarz, V. and Sutcliffe, K. W. (1959) Defective lactose
absorption causing malnutrition in infancy. Lancet,I, 1126.
7. Cook, G. Hypolactasia: geographical distribution, diagnosis, and practical
significance. In: Chandra, R. K. (ed.) Critical Reviews of Tropical Medi-
cine, 1984, pp. 117 – 139. Plenum Press, New York and London.
8. Brostoff, J. and Challacombe, S. J. (eds) (2002) Food allergy and intoler-
ance. Saunders, London, 2nd edition. ISBN 0702020389.
9. Brostoff, J. and Gamlin, L. (1989) The complete guide to food allergy and
intolerance. ISBN 0-7475-0242-0. Bloomsbury, London.
10. Bartoletus, F. (1633) Methodus in dyspnoeam seu de respirationibus libri
IV, quintus pro colophone accessit de curationibus ex dogmaticorum et
hermeticorum poenu depromtis. Bolgna, Boniniae, p. 400.
11. Plimmer, R. H. A. (1906 1907) On the presence of lactase in the intestine
of animals and on the adaptation of the intestine to lactose. J. Physiol., 35,
20 – 31.
12. Freund, J.-N., Gosse
´, F. and Raul, F. (1991) Derivatives of plant beta-
glucans are hydrolysed by intestinal lactase-phlorizin hydrolase of mam-
mals. Enzyme,45, 71– 74.
13. Mackey, A. D., Henderson, G. N. and Gregory, J. F. (2002) Enzymatic
hydrolysis of pyridoxine-5’-beta-D-glucoside is catalysed by intestinal
lactase-phlorizin hydrolase. J. Biol. Chem., 277, 26858– 26864. The molecular basis of lactose intolerance 199
Black plate (200,1)
14. Skovbjerg, H., Noren, O., Sjostrom, Danielsen, E. M. and Stig, B. (1982)
Further characterization of intestinal lactaseyphlorizin hydrolase. Bio-
chim. Biophys. Acta,707, 89– 97.
15. Ramaswamy, S. and Radhakrishn A. N. (1975) Lactase-phlorizin hydro-
lase complex from monkey small intestine: purification, properties and
evidence for two catalytic sites. Biochim. Biophys. Acta,403, 446 – 455.
16. Aribas, J. C. D., Herrero, A. G., Martin-Lomas, M., Canada, F. J., He, S.
and Withers, S. G. (2000) Differential mechanism-based labeling and
unequivocal activity assignment of the two active sites of intestinal
lactaseyphlorizin hydrolase. Eur. J. Biochem., 267, 6996– 7005.
17. Swallow, D. M. (2003) Genetics of lactase persistence and lactose intol-
erance. Ann. Rev. Genet., 37, 197– 219.
18. Troelsen, J. T., Mitchlemore, C., Spodsberg, N., Jensen, A. M. J., Noren,
O. and Sjostrom, H. (1997) Regulation of lactase-phlorizin hydrolase gene
expression by the caudal-related homeodomain protein cdx-2. Biochem. J.,
322, 833– 838.
19. Fang, R., Olds, L., Santiago, A. and Sibley, E. (2001) GATA family of
transcription factors activate lactase gene promoter in intestinal Caco-2
cell. Am. J. Physiol., 280, G58– G67.
20. Radebach, J. R., Wurthrich, M., Grunberg, J., Sterchi, E. E. and Naim,
H. Y. (1996) Maturation of human intestinal lactase-phlorizin hydrolase:
generation of the brush border form of the enzyme involves at least two
proteolytic cleavage steps. Eur. J. Biochem., 236, 789 – 795.
21. Mantei, N., Villa, M., Enzler, T., Wacke, H., Boll, W., James, P.,
Hunziker, W. and Semenza, G. (1988) Complete primary structure of
human and rabbit lactase-phlorizin hydrolase: implications for biosynth-
esis, membrane anchoring and evolution of the enzyme. EMBO J., 7,
2705 – 2713.
22. Wuthrich, M., Grunberg, J., Hahn, D., Jacob, R., Radebach, I., Naim, H.
Y. and Sterchi, E. E. (1996) Proteolytic processing of human lactase-
phlorizin hyrolase is a two-step event: indentification of the cleavage sites.
Arch. Biochem. Biophys., 336, 27– 34.
23. Waud, J., Bermudez-Fajardo, A., Sudadhaharan, T., Jefferies, J., Jones,
A. and Campbell, A. K. (2001) Measurement of proteases using chemi-
luminescence resonance energy transfer chimaeras between green fluor-
escent protein and aequorin. Biochem. J., 357, 687– 697.
24. Jacob, R., Peters, K. and Naim, H. Y. (2002) The prosequence of human
lactase-phlorizin hydrolase modulates the folding of the mature enzyme. J.
Biol. Chem., 277, 8217– 8225.
25. Maiuri, L., Rossi, M., Raia, V., Paparo, F., Garipoli, V. and Auricchio, S.
(1993) Surface staining on the villus of lactase protein and lactase activity
in adult-type hypolactasia. Gastroenterology,105, 708– 714.
26. Lifshitz, F. (1996) Congenital lactose deficiency. J. Pediatr., 69, 229– 237.
27. Kuokkanen, M., Kokkonen, J., Enattah, N. S., Ylisaukko-Oja, T., Koma,
H., Varilo, T., Peltonen, L., Savilahti, E. and Jarvela, I. (2006) Mutations
in the Translated Region of the Lactase Gene (LCT) Underlie Congenital
Lactase Deficiency. Am. J. Hum. Genet., 78, 339 – 44. Epub 2005.
28. Tivey, D. R., Morovat, A. and Dauncey, M. J. (1993) Administration of 3,
5, 3’-triiodothyronine induces a rapid increase in enterocyte lactase-
200 Anthony K. Campbell, Jonathan P. Waud and Stephanie B. Matthews
Black plate (201,1)
phlorizin hydrolase activity of young pigs on a low energy intake. Exp.
Physiol., 78, 337– 346.
29. Matosin-Matekalo, M., Meseron, J. E., Delezay, O., Poiree, J. C.,
Ilundain, A. A. and Brot-Laroche, E. (1998) Thyroid hormone regulation
of the Na
yglucose cotransporter SGLT1 in caco-2 cells. Biochem. J.,
334, 633– 640.
30. Srinivasan, D. and Minocha, A. (1998) When to suspect lactose in
tolerance. Symptomatic, ethnic, and laboratory clues. Postgraduate
Med., 104, 109– 123.
31. Llewellyn, D. H., Roderick, L., Sheik, N. and Campbell, A. K. (1996)
Calcium stress induces calreticulin gene expression. Biochem. J., 318, 400 –
32. Flatz, G. (1987) Genetics of lactose digestion in humans. Adv. Hum.
Genet., 16, 1– 77.
33. Enattah, N. S., Sahi, T., Savilahti, E. et al. (2002) Identification of a variant
associated with adult-type hypolactasia. Nature Genetics,30, 233 – 237.
34. Kuokkanen, M., Enattah, N. S., Oksanen, A., Savilahti, E., Orpana, A.
and Jarvela, I. (2003) Transcriptional regulation of the lactase phlorizin-
hydrolase gene by polmorphisms with adult-type hypolactasia. Gut,52,
647 – 652.
35. Campbell, A. K. (1994) Rubicon: the fifth dimension of biology. Duck-
worth, London. Now available from Welston Press on CD.
36. Sharpe, P. T. (2000) Homeobox genes in initiation and shape of teeth
during development in mammalian embryos. In: Teaford, M. F., Smith,
M. M. and Fergusin, F. W. (eds.) Development, function and evolution of
teeth, pp. 3 – 12. Cambridge University Press, Cambridge.
37. Sosulski, F. W., Elkowicz, L. and Reichart, R. D. (1982) Oligosaccharides
in eleven legumes and their classified protein and starch fractions. J. Food
Sci., 47, 498– 502.
38. Bock, A. and Sawers, G. (1996) Fermentation. In: Neidhardt, F. C. (ed.),
Escherichia coli and Salmonella: cellular and molecular biology. 2nd edn.
Vol 1, Chap. 18, pp. 262 – 282.
39. Harden, A. and Walpole, G. S. (1906) Chemical action of Bacillus lactose
aerogenes (Escheris) on glucose and mannitol: production of 2, 3-butyl-
glycerol and acetyl-methylcarbinol. Proc. R. Soc. Lond. SerB, 77, 399 –
40. Campbell, A. K. (1983) Intracellular calcium: its universal role as regulator,
pp. 556, Chichester, John Wiley and Sons.
41. Jones, H. E., Holland, I. B., Baker, H. L. and Campbell, A. K. (1999)
Slow changes in cytosolic free Ca
in Escherichia coli highlight two
putative influx mechanisms in response to changes in extracellular cal-
cium. Cell. Calcium,25, 265– 274.
42. Campbell, A. K., Matthews, S. B. and Wann, K. T. (2004) Lactose causes
heart arrhythmia in the water flea Daphnia pulex.Comp. Biochem.
Physiol., B 139, 225– 234.
43. Metchnikoff, E. (1900) Sur les cytotoxines. Annales Institute Pasteur xiv, 402.
44. Metchnikoff, E. (1908) The Nature of Man: Studies in optimistic philoso-
phy, pp. 73. English Translation Chalmers Mitchell, P. Putnam’s and Son,
New York and London. The molecular basis of lactose intolerance 201
Black plate (202,1)
45. Frissora, C. L. and Koch, K. L. (2005) Symptom overlap and comorbidity
of irritable bowel syndrome with other conditions. Curr. Gastroenterol.
Rep., Aug;7(4), 264 – 271.
46. Szilagyi, A., Malolepszy, P., Yesovitch, S., Nathwani, U., Vinokuroff, C.,
Cohen, A. and Xue, X. (2005) Inverse dose effect of pretest dietary lactose
intake on breath hydrogen results and symptoms in lactase nonpersistent
subjects. Dig. Dis. Sci., Nov;50(11), 2178 – 2182.
47. Campbell, A. K. and Matthews, S. B. (2005) Darwin’s illness revealed.
Post. Grad. Med. J., 81, 248– 251.
48. Campbell, A. K. (2003) What Darwin missed. Astrophys. Space Sci., 285,
49. Campbell, A. K. (2003) Save those molecules! Molecular biodiversity and
life. J. Appl. Ecol., 40, 193– 203.
50. Simoons, F. J. (1971) The antiquity of dairying in Asia and Africa. Geogr.
Rev., 61, 431– 439.
51. Feldman, M. W. and Cavalli-Sforza, L. L. (1989) On the theory of
evolution under genetic and cultural transmission with application to
the lactose absorption problem, Chap. 8, pp. 145 – 173. In: Feldman, M.
W. (ed.), Mathematical Evolutionary Theory. Princeton University Press,
52. Heilskow, N. S. C. (1951) Studies on animal lactase. II. Distribution in
some of the glands of the digestive tract. Acta Physiol. Scand., 24, 84– 89.
53. Elwood, P. C. (2005) Milk and cardiovascular disease: a review of the
evidence. Am. J. Dairy Technol., 59, 502– 505.
54. Moss, M. and Freed, D. (2002) The cow and the coronary: epidemiology,
biochemistry and immunology. Int. J. Cardiol., 87, 203 – 216.
55. Williams, R. (1985) Biochemical individuality: the basis of the genetotrophic
concept. Keats Publishing, Connecticut.
1., Conventional information about lactose intolerance
and selling products.
2., Conventional information about lactose
intolerance and milk allergies, with symptoms, treatments, and ‘hidden’
milk products.
3., Conventional information only about lactose
intolerance and milk allergies.
4., The only one at present that deals with systemic
symptoms caused by lactose
202 Anthony K. Campbell, Jonathan P. Waud and Stephanie B. Matthews
... Only lactose provides 40% of the energy needs of suckling mammals. This fact explains why almost all the mammalian milk contains 40-75 g of lactose per litter, and why the milk of mammals is the only source in nature with a significant content of lactose [1][2][3]. Congenital deficiency to digest lactose is rare in baby mammals since it can lead to growth delay, dehydration, and even the death [3]. ...
... This system comprises the enzyme galactosyl transferase and the protein modifier α-lactalbumin. When the protein modifier binds to the galactosyl transferase, it catalyzes the synthesis of lactose from uridine-diphosphate-galactose (UDP-gal) and glucose [1][2][3]. In the absence of the protein modifier, the galactosyl transferase does not synthesize lactose and instead catalyzes the synthesis of N-acetyl lactosamine on glycoproteins. ...
... This last reaction occurs in most tissues, but in the mammary gland of women after giving birth, the increase in prolactin and a decrease in progesterone hormones induce the formation of the protein modifier (α-lactalbumin). Consequently, the breast can synthesize lactose in the milk for the nourishment of newborn mammals [2,3]. ...
... Sucrose is a fructose and glucose bonded by an α-1,4 glycosidic bond (Campbell et al., 2005). Sucrose can be included in dairy cow diets as a source of readily fermentable energy. ...
... Similar to sucrose, lactose can be included in dairy cow diets. Lactose is a galactose and glucose connected by a β 1,4-glycosidic bond (Campbell et al., 2005). Lactose comes from milk; it is synthesized in the mammary gland of cows using two glucose molecules. ...
Full-text available
Carbohydrates are one of the three macronutrients that provides energy in diets and are classified by their structures. Starch is a nonstructural carbohydrate and polysaccharide made of glucose monomers used for storage in plant cells. When starch makes up greater than 30% of the DM in diets there can be adverse effects on NDF digestibility due to decreases in ruminal pH. Sugars are water soluble carbohydrates that consist of monosaccharide and disaccharide units. Sugars ferment faster than starch because microorganisms in the rumen can ferment carbohydrates at different rates depending on their structure; however, this has not been shown to have negative effects on the ruminal pH. Sources of sugars such as molasses (sucrose) or whey (lactose) can be included in the diet as a partial replacement for starch in dairy cow diets. The purpose of replacing starch with sugars in a diet would be to add differing sources of carbohydrates in the diet to allow for continual fermentation of carbohydrates by the microorganisms in the rumen. It has been seen in studies and previous literature that the partial replacement of starch with sugars has the potential to maintain the ruminal environment and milk yield and composition in dairy cows without reducing NDF digestibility. The objective of this review is to evaluate the effects of partially replacing starch with sugars in dairy diets and its implication on ruminal fermentation, nutrient utilization, milk production, and feeding replacement strategy.
... Lactose sugar (b-galactosyl-1,4 glucose) is the main carbohydrate found in milk and human milk constitutes around 7% and 4.8% in bovine milk by weight of total milk (12). It is hydrolyzed to glucose and galactose by lactase enzyme produced in the lining of the small intestines in the human body (13,14).The amount of lactose sugar varies highly among various types of milk and also among dairy products as shown in Table 1. Lactase enzyme is a bgalactosidase (EC ...
Milk is the most common food consumed worldwide and is also a major ingredient in the preparation of various dairy products. However, despite the high production and consumption of milk and milk-based products, there is a large percent of the world’s population that suffer from allergies to milk solids and lactose intolerance. Lactose intolerance specifically means the inability of the body to breakdown the sugar to its simplest form for assimilation and it is due to the inefficiency or lack of the enzyme in the human body. The most convenient prevention method for the affected population is to avoid milk and milk-based products but this may be a cause of development of other health related issues that result from inadequate nutrient consumption. To help find an alternative to this problem, this study aims at first studying the underlying information on lactose intolerance and then studying plant-based beverages as a possible alternative to milk and milk-based products. • Key teaching points • Lactose intolerance specifically means the inability of the body to breakdown the sugar to its simplest form for assimilation and it is due to the inefficiency or lack of the enzyme in the human body. • Consumption of probiotics may help relieve the symptoms of lactose intolerance. • Soy beverage can be an economical alternative for lactose intolerant populations and has calcium content comparable to bovine milk. • Calcium absorption in fortified plant based beverages depends upon type of calcium salt used.
... Galaktoza se nešto efikasnije apsorbuje od glukoze. Nakon apsorpcije, 94% galaktoze u sklopu Leirovog puta biosinteze u jetri prelazi u glukozu, dok se ostatak galaktoze metaboliše u eritrocitima ili izbacuje urinom (1,4). Laktazna aktivnost raste tokom zadnje trećine intrauterinog razvoja i kod svih sisara održava visok nivo sve do odbijanja od dojke. ...
Full-text available
Lactose is a disaccharide found in milk and dairy products. Children and adults with lactose intolerance are unable to tolerate significant amounts of lactose because of an inadequate amount of the enzyme lactase. The condition occurs in three main types: primary, secondary, and primary adult-type hypolactasia. The use of milk in the diet of these individuals may lead to appearance of the irritable bowel syndrome. In persons with lactose intolerance symptoms include diarrhoea, dominated by abdominal colic, loud peristaltic sounds, increased flatulence and meteorism. A diagnosis of lactose intolerance can usually be made with a careful history, elimination of lactose from the diet, lactose tolerance test, hydrogen breath test and genetic testing. In the absence of appropriate tests in patients with suspected primary adult-type hypolactasia, diagnosis can be made as in patients with food allergy. Treatment is based on the restriction of lactose intake with the use of fermented milk products. However, especially for children, if milk and dairy products are eliminated from the diet, it is important to ensure D vitamin and calcium supplementation.
... However, LI has been recognized and diagnosed as a medically important disease only in the past 50 years (4,5) . LI is a pathological condition characterized by the inability to digest lactose due to the absence or insufficient activity of lactase enzyme (β-galactosidase) (6,7). ...