Hypoglycemic effect of lupin seed γ-conglutin in experimental animals and healthy human subjects

Abstract

A lupin seed γ-conglutin-enriched preparation was tested in a glucose overload trial with both murine models and adult healthy volunteers. The results with rats showed a dose-dependent significant decrease of blood glucose concentration, which confirmed previous findings obtained with the purified protein. Moreover, three test-product doses equivalent to 630, 315, and 157.5 mg γ-conglutin, orally administered 30 min before the carbohydrate supply, showed a relevant hypoglycemic effect in human trials. Insulin concentrations were not significantly affected. The general hematic parameters did not change at all. This is the first report on the glucose-lowering effect of lupin γ-conglutin in human subjects.

Full text

1Hypoglycemic effect of lupin seed γ-conglutin in experimental animals and
2healthy human subjects
3Juan C. Bertoglio
a
, Mario A. Calvo
a
, Juan L. Hancke
b
, Rafael A. Burgos
b
, Antonella Riva
c
,
4Paolo Morazzoni
c
, Cesare Ponzone
c
, Chiara Magni
d
, Marcello Duranti
d,
⁎
5
a
Instituto de Medicina, Valdivia, Chile
6
b
Instituto de Farmacologia Universidad Austral de Chile,Valdivia, Chile
7
c
Indena S.p.A., Milan, Italy
8
d
Dept. of AgriFood Molecular Sciences, Università degli Studi di Milano, Italy
9
11 article info12 abstract
13 Article history:
14 Received 18 March 2011
15 Received in revised form 6 May 2011
16 Accepted 8 May 2011
17 Available online xxxx
18 A lupin seed γ-conglutin-enriched preparation was tested in a glucose overload trial with both
19 murine models and adult healthy volunteers. The results with rats showed a dose-dependent
20 significant decrease of blood glucose concentration, which confirmed previous findings
21 obtained with the purified protein. Moreover, three test-product doses equivalent to 630, 315,
22 and 157.5 mg γ-conglutin, orally administered 30 min before the carbohydrate supply, showed
23 a relevant hypoglycemic effect in human trials. Insulin concentrations were not significantly
24 affected. The general hematic parameters did not change at all.
25 This is the first report on the glucose-lowering effect of lupin γ-conglutin in human subjects.
26 © 2011 Published by Elsevier B.V.
27 Keywords:
28 Antidiabetic
29 Lupinus albus
30 Protein extract
31 Animal trial
32 Human study3334
35
36 Introduction
37 Hyperglycemia is recognized to be the central feature of all
38 unbalances in the metabolism of carbohydrates, lipids,
39 ketones and amino acids [1]. The most diffused pathological
40 condition, characterized by stable hyperglycemia, is known
41 today as type-2 diabetes (accounting for about 90% of all
42 diabetes cases). Nowadays, type-2 diabetes is considered an
43 epidemic disease, especially in the Western countries, where
44 the incidence in the population is estimated to range from 2%
45 up to 4% [2]. Type-2 diabetes is usually preceded by years of
46 an abnormal condition, termed impaired glucose tolerance
47 (IGT) characterized by plasma glucose levels between 140
48and 199 mg/dL, 2 hours after a standard oral glucose
49challenge, but not as high as in diabetes (N200 mg/dL).
50Several characteristics in the population have been recog-
51nized to be associated with a greater risk of progression from
52IGT to type 2 diabetes. Among these are impaired insulin
53secretion, insulin resistance, obesity and age Q1[2-4]. Therefore,
54actions aimed at controlling this situation are crucial.
55Lupin is a leguminous seed which has largely been used as
56a food for its high protein content in the Mediterranean area
57since thousand years. Around the year 1000 AD, the Persian
58doctor Ibn Sīnā,amongthefirsts to describe diabetes
59symptoms, used mixtures of lupin, fenugreek and zedoary
60seed flours to remarkably reduce sugar excretion. Lupin seed
61is mentioned in the ancient and traditional pharmacopoeia
62books as an anti-diabetic product. Last century, in the search
63of the lupin seed active principle, Orestano described the
64extraction and purification from white lupin seeds of a
65compound capable of decreasing glycemia in rabbits. How-
66ever, the extraction was tedious and the yield extremely low
67and thus the product was considered to be lacking of any
Fitoterapia xxx (2011) xxx–xxx
Abbreviations: BMI, body mass index; BUN, blood urea nitrogen; GOT,
glutamate-oxaloacetate transaminase; GPT, glutamate-pyruvate transami-
nase; GGT, γ-glutamyl transpeptidase; LDH, lactate dehydrogenase.
⁎Corresponding author at: Department of AgriFood Molecular Sciences,
Università degli Studi di Milano, Via Celoria, 2 I-20133 Milano, Italy.
Tel.: +39 02 503 16817; fax: + 39 02 503 16801.
E-mail address: marcello.duranti@unimi.it (M. Duranti).
FITOTE-02218; No of Pages 6
0367-326X/$ –see front matter © 2011 Published by Elsevier B.V.
doi:10.1016/j.fitote.2011.05.007
Contents lists available at ScienceDirect
Fitoterapia
journal homepage: www.elsevier.com/locate/fitote
Please cite this article as: Bertoglio JC, et al., Hypoglycemic effect of lupin seed γ-conglutin in experimental animals and
healthy human subjects, Fitoterapia (2011), doi:10.1016/j.fitote.2011.05.007

68 potential application [5]. The possibility that a lupin seed
69 protein could be the active principle was not taken in
70 consideration until recently.
71 During our research activities on lupin seed proteins, we
72 undertook a series of research activities aimed at unraveling the
73 specific hypoglycemic role of one of the lupin proteins, termed γ-
74 conglutin. Lupin γ-conglutin is a homo-tetrameric glycoprotein
75 in which the monomeric unit consists of two disulphide linked
76 heterogenous subunits of about 30 and 17 kDa (for a review on
77 γ-conglutin properties, see reference 6). From the viewpoint of
78 γ-conglutin biological activity, orally-administered pure γ-
79 conglutin was found to effectively decrease plasma glucose in
80 glucose overloaded rats in a dose-dependent manner [7].More
81 recently, the insulin-mimetic activity of γ-conglutin on differen-
82 tiating myoblasts was described [8].
83 In this work, the remarkable glucose lowering capacity of
84 a lupin γ-conglutin-enriched preparation was confirmed in
85 rat models and assessed for the first time in healthy human
86 subjects.
87 2. Materials and methods
88 2.1. Lupin seed γ-conglutin laboratory purification
89 Lupinus albus L. seed γ-conglutin was purified to homogene-
90 ity from dehulled dry lupin seeds by using a combination of
91 chromatographic steps, as described by Duranti et al. [9].
92 2.2. Lupin seed γ-conglutin preparation for animal and human
93 studies
94 A lupin γ-conglutin-enriched dry extract (Pro-Gamma
™
)
95 was prepared by INDENA S.p.A., Milano, Italy, according to an
96 industrially-developed procedure (WO 2004/071521) which
97 consisted essentially in a protein wet extraction process and
98 solvent precipitation. The test product for the study was
99 manufactured in the form of dried powder. The protein
100 content of the dry powder, as assessed by the Lowry method
101 [10], was 44.8% of the dry weight. γ-Conglutin content,
102 evaluated by SDS-PAGE densitometric scanning, was about
103 47% of the total proteins in the dry powder (see below).
104 The test product was used as such in rat studies at the
105 dosages of 50, 100 and 200 mg/kg b.w. (see below details on
106 the administration procedure). As far as the human trials are
107 concerned, the dry powder was packed in vacuum sealed
108 sachets, for easy handling, containing 1500 mg powder each
109 (equivalent to 315 mg of γ-conglutin) by FARMINDUSTRIA
110 S.A. Laboratories (Santiago, Chile). For placebo, the formula
111 was the same as for the verum formulation, except for the
112 exclusion of lupin dry extract, which was substituted by
113 microcrystalline cellulose (Avicel PH302). The test product
114 and the placebo were kept at room temperature in a dry place
115 and protected from light.
116 2.3. SDS-PAGE
117 SDS-PAGE was carried out on 12% polyacrylamide gels,
118 according to Laemmli [11] under reducing and non-reducing
119 conditions using a mini-PROTEAN II cell (Bio-Rad). The gels
120 were Coomassie blue stained.
121The SDS-PAGE images were acquired using a scanner,
122Canoscan 8000F (Canon, Milan, Italy) interfaced with a
123personal computer and densitometric scans were carried
124out with ImageMaster 1D software (Amersham Pharmacia
125Biotech, Milan, Italy).
1262.4. Design of the animal study
127A total of 100 male rats (Charles River, Calco, LC, Italy)
128with an average body weight ranging between 275 and 300 g
129were maintained under stable conditions for 7 days before
130the experiment. The animals were given a standard rat diet
131and were kept under automatically controlled light, temper-
132ature, and humidity conditions. The rats were divided into
133five groups. One group, the control group, received only the
134control product, i.e., the placebo; three groups received 50,
135100, or 200 mg/kg body weight of the γ-conglutin-enriched
136test product, corresponding to 10.5, 21.0 and 42.0 mg γ-
137conglutin; the last group was given 50 mg/kg body weight of
138metformin added to the control product. Administration was
139carried out by gavage, 30 min before the glucose overloading
140experiment. At time 0 of the experiment, each rat was given
1412 g/kg body weight glucose administered orally. At estab-
142lished times thereafter (0, 30, 60 and 90 min from glucose
143administration), each rat in a group of 5 rats per time and
144dose was treated with 50 mg/kg body weight Na-thiopenthal,
145and 5 mL of blood were collected in 7.5 mmol/L EDTA
146containing tubes. The blood was immediately centrifuged at
1472000×gat 4 °C for 10 min and the supernatant used for
148glucose assays. All procedures involving rats and their care
149were performed according to the Italian Government Guide-
150lines for animal tests and were in agreement with the
151European Commission rules (86/609/EEC).
1522.5. Design of the human trial
153The present trial was a placebo-controlled study con-
154ducted on fifteen adult healthy volunteers, who received
155three different single test doses of respectively 750 (half the
156content of 1 sachet), 1500 (1 sachet) and 3000 mg (2 sachet)
157of the test product, corresponding to 157.5, 315 and 630 mg
158γ-conglutin, respectively, and a placebo, with no γ-conglutin
159added. The whole duration of the trial was 7 weeks. The three
160doses and the placebo were administered per os 30 min
161before a carbohydrate meal, consisting in one serving of 85 g
162boiled (Grade 1) white rice which corresponded to an intake
163of 75 g carbohydrate. The flowchart of the clinical study is
164reported in Table 1.
165The study was organized and directed by Patagonia
166Clinical Trials, CRO program of the Institute of Pharmacology,
167Universidad Austral de Chile in the city of Valdivia, Chile. All
168the volunteers were recruited from the city of Valdivia. The
169study started after approval by the local Bioethical Committee
170with the recruitment of the volunteers and was completed
171within 5 weeks from the beginning of the first session.
172Written informed consent was obtained from all subjects.
1732.6. Recruitment of the subjects
174Fifteen healthy adult (N18 years old) male and female
175volunteers, who complied with the inclusion criteria, were
2J.C. Bertoglio et al. / Fitoterapia xxx (2011) xxx–xxx
Please cite this article as: Bertoglio JC, et al., Hypoglycemic effect of lupin seed γ-conglutin in experimental animals and
healthy human subjects, Fitoterapia (2011), doi:10.1016/j.fitote.2011.05.007

176 selected in November 2008. The selected adult volunteers were
177 not receiving any acute or chronicpharmacological therapy and
178 had a body mass index (BMI) less than 30 kg/m
2
. All subjects
179 who were enrolled for the present study had a fasting plasma
180 glucose (BG) b100 mg/100 mL (or b5.6 mmol/L), and a normal
181 oral glucose tolerance test (OGTT), defined as a plasma glucose
182 between 100 mg/100 mL and 140 mg/100 mL (or 5.6 and
183 7.8 mmol/L, respectively), at 120 min after a 75 g serving of
184 carbohydrates [2,4,12].
185 2.7. Exclusion criteria
186 The following criteria were used to exclude candidate
187 volunteers:
188 –elevation of glycemia in plasma up to 140 mg/dL or
189 above, after 75 g of carbohydrate load;
190 –diabetes mellitus;
191 –hospitalization over the last 60 days;
192 –arrhythmia.
193 –renal failure (blood creatinine N1.5 mg/dL);
194 –documented allergy to peanuts;
195 –any other chronic or limiting disease, including
196 alcoholism;
197 –any chronic pharmacological treatment;
198 –any disease or condition that the enrolling physician
199 considered making the subject unsuitable for participation
200 to this trial;
201 –pregnant women and women under hormonal contra-
202 ception method prescribed less than 2 months before.
203 2.8. Statistical analysis
204 Results were depicted as the mean ±SME. The incremen-
205 tal area under the glucose and insulin curves were calculated
206 as described by Wolever and Jenkins [13]. The statistical
207 significance was calculated using the paired t-test and a
208 statistical significance of Pb0.05 was adopted.
209 In addition, an exploratory analysis for all dataset and
210 separately for each treatment group was also performed.
211 Descriptive statistics and 95% confidence intervals were used
212 to estimate parameters. To evaluate normality assumption we
213 used Shapiro–Wilks test, considering non-normal distribution
214 those presenting p-values lower than 0.05. To evaluate the
215 degree of independence between repeated measures during
216 the period of study, the Pearson correlation test was used. A
217moderate to high intra-individual correlation to all outcome
218variables over repeated measurements was observed (data not
219shown),therefore a generalizedestimating equation(GEE) was
220used to evaluate these data, considering an exchangeable
221correlation structure by robust standard errors, using the
222identity link function which works with normal distribution
223[14-17]. The outcome variables (glucose and insulin) were
224modeled considering treatment variable [(0=placebo; dose 1
225(750 mg), 2 (1500 mg) and 3 (3000 mg)] plus interaction
226variable (group*time). This model was adjusted by body max
227index (BMI). In all adjusted models, the interaction terms were
228non-significant, indicating that the treatment effect on the
229outcome does not vary significantly over the time among
230groups (data not shown). The xtgee procedure of STATA was
231used to evaluate these models [18]. All analyses were done
232considering on a base of intention to treat. Graphical methods
233were used to show glucose and insulin changes over time
234(minutes). In all models, timevariable was used as a continuous
235variable.
2363. Results and discussion
2373.1. Characterization of γ-conglutin preparations by SDS-PAGE
238The electrophoretic pattern under reducing and non-
239reducing conditions of the sample product used in this work
240for the animal and human studies (Fig. 1, lanes 2) is compared
241with the placebo preparation, which did not contain γ-
242conglutin (Fig. 1, lanes 1) and a laboratory purified γ-conglutin
243control (Fig. 1, lanes 3). The characteristic main band of γ-
244conglutin under non-reducing conditions with anM
r
of about
24550 kDa was visible in the panel B, lanes 2 and 3. When reducing
246conditions were applied, the50 kDa polypeptide of γ-conglutin
247underwent reduction giving rise to the 30 and 17 kDa subunits
248of this lupin protein [6], which can be seen in the panel A of the
249figure.
250As far as purity of the sample is concerned, a number of
251minor bands were visible in the industrial γ-conglutin
252preparation, but not in the laboratory purified protein, as
253expected. Therefore, the relative amounts of γ-conglutin
254polypeptides in the industrial sample were quantified by
255densitometric scanning of both reduced and non-reduced
256samples in order to assess the actual γ-conglutin amounts in
257the formulates. With the reduced sample, 46.9± 5.0% γ-
258conglutin with respect to the total polypeptides was obtained
259in triplicate analyses, while 60.0 ± 7.0% γ-conglutin was
260obtained with the non-reduced sample. This difference can
Table 1t1:1
Human trial flowchart of each experimental session.
t1:2
t1:3Time (minutes) BMI Plasma glucose and insulin Creatinine, BUN, total cholesterol, triglycerides,
GOT, GPT, GGT, bilirubin, alkaline phosphatase, LDH.
t1:4−10 (Baseline) Yes Yes In sessions: placebo, 750 and 3000 mg, only
t1:50 (Test product) No No No
t1:615 No Yes No
t1:730 (Carbohydrate intake) No Yes No
t1:860 No Yes No
t1:990 No Yes No
t1:10 120 No Yes No
t1:11 180 No Yes In sessions: placebo, 750 and 3000 mg, only
3J.C. Bertoglio et al. / Fitoterapia xxx (2011) xxx–xxx
Please cite this article as: Bertoglio JC, et al., Hypoglycemic effect of lupin seed γ-conglutin in experimental animals and
healthy human subjects, Fitoterapia (2011), doi:10.1016/j.fitote.2011.05.007

261 be explained with the presence of insoluble protein material
262 in the non-reduced sample, which was removed before the
263 electrophoretic run or did not enter the gel network. On this
264 basis, the lower γ-conglutin estimate, i.e., 46.9% of the total
265 proteins, and the protein content of the dry powder, i.e.,
266 44.8%, were both considered for the quantification of γ-
267 conglutin as the active principle concentration in the
268 formulates. According to this evaluation, the amounts of γ-
269 conglutin in the three administered dosages of formulates
270 used for human studies, i.e., 750, 1500 and 3000 mg, were
271 157.5, 315 and 630 mg, respectively.
272 3.2. Results of the animal study
273 The γ-conglutin-enriched preparation was orally admin-
274 istered to rats prior to glucose overload and blood glucose
275 concentrations were monitored for 90 min. The whole
276 procedure is detailed under Materials and methods. The
277 administration of 2 g/kg glucose caused a fourfold increase of
278 blood glucose in control rats from 70 ±9 to 285 ±28 mg/dL;
279 Pb0.001. The greatest increase occurred after 30 min from
280 glucose intake and then, as physiologically expected, concen-
281 trations decreased (not shown). Rat pre-treatment with
282 the test product administered prior to glucose overload at
283 the dosages of 50, 100 and 200 mg/kg b.w (corresponding
284 to 10.5, 21.0 and 42.0 mg γ-conglutin) decreased the
285 blood glucose increase in a dose-dependent pattern. As it
286 can be seen in Fig. 2, the areas under the curve (AUC) of
287 glucose concentrations decreased by 14% (PN0.05), 42%
288 (Pb0.01) and 64% (Pb0.001) at the three mentioned doses,
289 respectively, in comparison to the control treatment. The
290 effect obtained with the highest dose did not significantly
291 differ from the action of metformin at 50 mg/kg b.w. This
292 finding confirms previous results obtained with pure γ-
293 conglutin [7], where the hypoglycemic effect of the lupin
294 protein was first described.
2953.3. Results of the clinical study
296The tolerability of the L. albus purified dry extract was fully
297satisfactory. No clinical subjective side effects were observed,
298nor any adverse events recorded upon administration and
299during the time course of this study. All vital signs remained
300unwavering, including appetite and fluid intake. Hemody-
301namic parameters were stable; clinical laboratory analysis of
302blood chemistry concerning creatinine, urea nitrogen (BUN),
303total cholesterol, triglycerides, glutamate-oxalacetate trans-
304aminase (GOT), glutamate-pyruvate transaminase (GPT),
305gamma-glutamil transpeptidase (GGT), bilirubin, alkaline
306phosphatase and lactate dehydrogenase (LDH), also
307remained unchanged and stable throughout (not shown).
308The glucose and insulin plasma levels (means ±SME)
309during the time course of the clinical study are shown in
AB
67
94
45
30
20
14
Mr,
kDa
11 2 23 3 2 21 1
Fig. 1. SDS-PAGE under reducing (A) and non-reducing conditions (B) of γ-conglutin preparations. The gels were stained by Coomassie brilliant blue. The lanes
with same numbers refer to two differently loaded volumes of the same sample. Sample lanes are the following: 1: placebo preparations without γ-conglutin; 2:
industrial scale γ-conglutin preparation; 3: laboratory scale γ-conglutin preparation (4 μg protein).
Dosage groups (mg test product/kg b.w.)
AUC (mg / dL x min90)
**
15000
10000
5000
0
Control 50 100 200 Metformin
*** ***
*
Fig. 2. Histogram of the areas under the curve, AUC, of plasma glucose
concentrations in the rat study during the 90 min trial at the three doses
indicated in the figure. No γ-conglutin was administered in the control group,
while metformin at 50 mg/kg b.w. was used as a positive control.The data are
means± SEM of five animals in each dosagegroup at 0, 30, 60 and 90 min from
glucose overload. *Pb0.05, **Pb0.01 and ***Pb0.001 vs. the control.
4J.C. Bertoglio et al. / Fitoterapia xxx (2011) xxx–xxx
Please cite this article as: Bertoglio JC, et al., Hypoglycemic effect of lupin seed γ-conglutin in experimental animals and
healthy human subjects, Fitoterapia (2011), doi:10.1016/j.fitote.2011.05.007

310 Fig. 3, panels A and B, respectively. The changes of glucose
311 concentrations in the plasma of all fifteen subjects at the three
312 doses tested, with respect to placebo, were remarkable. In
313 particular, statistically significant and comparable reduction
314 with the three doses of formulate was observed at 60 min,
315 where the average glucose concentration was around 81% of
316 the placebo control. At 90 min, the shapes of the curves for
317 the three doses differentiated, being the lowest dose more
318 similar to the placebo control (83%). Conversely, the
319 intermediate dose, 1500 mg, had the greatest effect (61% of
320 the placebo) and the greatest dose, 3000 mg, resulted in 71%
321 of the placebo. At 120 min, the highest doses still maintained
322 a statistically significant reduction, being the 71% of the
323 placebo control. The overall shapes of the highest doses
324 curves were very similar, thus suggesting that the 1500 mg
325 dose already reached its maximum effect.
326 In parallel, insulin plasma concentrations of all fifteen
327 subjects, shown in panel B of Fig. 3, showed an increase in the
328 absolute amount of secreted insulin at 60 min at all doses,
329then a relatively homogenous pattern of decay followed, with
330the data not significantly differing from the placebo.
331Therefore, the main outcome of these results was the
332efficacy of γ-conglutin oral administration on the modifica-
333tion of post-prandial plasma glucose, while the overall
334pattern of insulin response was less affected. The hypoglyce-
335mic effect of laboratory-purified homogenous γ-conglutin
336had been already observed in glucose overloaded animal
337models [7] and its bioactivity was demonstrated in cell
338models too [8]. Therefore, the identification of γ-conglutin as
339the active principle is not under discussion in this work.
340Data are also reported as variations of the areas under the
341curve (AUC) in Fig. 4, panels A and B for glucose and insulin,
342respectively. Glucose concentration decreased in a statistically
343significant manner at the two highest doses of γ-conglutin, 75%
344and 79% of the placebo for 1500 mg and 3000 mg, respectively.
345The lowest dose also showed a decreasing trend, but its
346difference with the placebo was not statistically significant. As
347far as insulin plasma levels are concerned, no statistically
348significant difference was observed upon the treatments.
349The subsequent statistical analysis by means of the GEE
350(adjusting glucose and insulin plasma levels for BMI and time)
351is reported in Table 2. The results showed that treatment vs.
352placebo variable did havea negative significant association only
A
030 60 90 120 150 180
50
75
100
125
150
Placebo
1,500 mg 3,000 mg
750 mg
time (min)
Glycaemia (mg / dL)
B
0
10
20
30
40
Placebo
1,500 mg 3,000 mg
750 mg
Insulin (µlU / mL)
0 30 60 90 120 150 180
time (min)
Fig. 3. Time course of plasma glucose (panel A) and insulin (panel B)
concentrations in healthy volunteers (n= 15) treated with the test product
and a placebo, after starch intake.
A
0 (Placebo) 750 1,500
3,000
0
2500
5000
7500
10000
** *
Lupin dosage groups (mg)
AUC (mg / dL x min180)
B
0 (Placebo) 750 1,500 3,000
0
1000
2000
3000
4000
Lupin dosage groups (mg)
AUC (µIU / mL x min180)
Fig. 4. Histogramof the areas under thecurve, AUC, of plasma glucose(panel A)
and insulin concentrations (panel B) in the human study during the 180 min
trial at the threedoses indicated in the figure.No γ-conglutin was administered
in the control group (placebo). Thedata are means ±SEM of 15 blood samples.
*Pb0.05, **Pb0.01 vs. the control.
5J.C. Bertoglio et al. / Fitoterapia xxx (2011) xxx–xxx
Please cite this article as: Bertoglio JC, et al., Hypoglycemic effect of lupin seed γ-conglutin in experimental animals and
healthy human subjects, Fitoterapia (2011), doi:10.1016/j.fitote.2011.05.007

353 with blood glucose levels. The three dosages used showed an
354 average decrease of −8.71 (−12.42 to −4.99; Pb0.001),
355 −11.35 (−17.23 to −5.46; Pb0.002) and −8.98 (−14.53 to
356 −3.43; Pb0.001) mg/dL, respectively. Conversely, no signifi-
357 cant effects on insulin plasma levels were observed.
358 Altogether, these results support the conclusion that γ-
359 conglutin does not affect insulin secretion, rather, by
360 decreasing glucose concentrations, plays a role as a insulin-
361 mimetic compound, as already suggested in a previous work
362 with cell models. Indeed, the treatment of myocites with γ-
363 conglutin was found to activate several proteins and enzymes
364 involved in the insulin signaling pathway, to trigger protein
365 biosynthetic processes and induce differentiating mecha-
366 nisms closely resembling those monitored with similar
367 insulin concentrations [8].
368 4. Conclusions
369 In this study a γ-conglutin-enriched preparation was orally
370 administered to rats and found to be as effective as in a previous
371 work [7], where a homogenous γ-conglutin laboratory prepara-
372 tion was analyzed, thus confirming the biological activity of this
373 lupin protein as a glucose-lowering agent. Moreover, in the
374 present work the same test product, orally administered to
375 healthy adults with good tolerability, proved to induce a
376 significant reduction of post-prandial plasma glucose concentra-
377 tion upon a standard load of carbohydrates, without a quantita-
378 tively significant modification of the insulin secretion response.
379 Even the lowest product dose of 750 mg (equivalent to 157.5 mg
380 γ-conglutin), significantly reduced the plasma glucose level
381 when adjusted for BMI. This seems particularly relevant in the
382 perspectives of using the extract in overweight subjects.
383 Several aspects, especially those related to γ-conglutin
384 metabolic fate and mechanism of action, still remain to be
385 investigated. Interestingly, recent data reporting stability of γ-
386 conglutin to digestive proteolytic attack, unless the protein is
387 completely denatured [19] would support the oral utilization of
388 this protein. Moreover, the possibility of transit of the lupin
389 protein through Caco-2 cell monolayers and in ex vivo models of
390 everted intestinal sacs has very recently been demonstrated [20].
391 With this work we have demonstrated for the first time that
392 the active protein responsible for the claimed anti-diabetic
393 effect of the lupin seed is effective in man, in addition to animal
394 models. Further studies are needed for the best exploitation of
395 this dietary protein in the food and nutraceutical areas.
396397398399400401402403404405406407408References
409[1] Goodman Gilman A. In: Hardman JG, Limbird LE, Goodman Gilman A,
410editors. The Pharmacological Basis of Therapeutics. 10th ed. New York:
411McGraw-Hill, Medical Publishing Division; 2001.
412[2] American Diabetes Association. Diagnosis and classification of diabetes
413mellitus. Diabetes Care 2004;27:s5–s10.
414[3] Gabir MM, Hanson RL, Dabelea D, Imperatore G, Roumain J, Bennett PH,
415et al. The 1997 American Diabetes Association and 1999 World Health
416Organization criteria for hyperglycemia in the diagnosis and prediction
417of diabetes. Diabetes Care 2000;23:1108–12.
418[4] The Expert Committee on the Diagnosis and Classification of Diabetes
419Mellitus. Follow-up report on the diagnosis of diabetes mellitus.
420Diabetes Care 2004;6:3160–7.
421[5] Orestano G. Sull'azione ipoglicemica dei semi di Lupinus albus. Arch
422Farmacol Sperim 1940;70:113–7.
423[6] Duranti M, Consonni A, Magni C, Sessa F, Scarafoni A. The major proteins
424of lupin seed: characterisation and molecular properties for use as
425functional and nutraceutical ingredients. Trends Food Sci Techn
4262008;19:624–33.
427[7] Magni C, Sessa F, Accardo E, Vanoni M, Morazzoni P, Scarafoni A, Duranti
428M. Conglutin γ, a lupin seed protein, binds insulin in vitro and reduces
429plasma glucose levels of hyperglycemic rats. J Nutr Biochem 2004;15:
430646–50.
431[8] Terruzzi I, Senesi P, Magni C, Montesano A, Scarafoni A, Luzi L, Duranti
432M. Insulin-mimetic action of conglutin-γ, a lupin seed protein, in mouse
433myoblasts. Nutr Metab Cardiovasc Dis 2011;21:197–205.
434[9] Duranti M, Scarafoni A, Gius C, Faoro F, Negri A. Heat induced synthesis
435and tunicamycin-secretion of the putative storage glycoprotein con-
436glutin γfrom mature lupin seeds. Eur J Biochem 1994;222:387–93.
437[10] Lowry OH, Rosebrough NJ, Farr AL, Randall RJ. Protein measurement
438with the Folin–Phenol reagents. J Biol Chem 1951;193:265–75.
439[11] Laemmli UK. Cleavage of structural proteins during the assembly of the
440head of bacteriophage T4. Nature 1970;227:660–5.
441[12] Dekker JM, Balkau B. Counterpoint: impaired fasting glucose: the case
442against the new American Diabetes Association guidelines. Diabetes
443Care 2006;29:1173–5.
444[13] Wolever TMS, Jenkins DJA. The use of the glycemic index in predicting
445the blood glucose response to mixed meals. Am J Clin Nutr 1986;3:
446167–72.
447[14] Mayer-Hamblett N, Kronmal RA. Improving the estimation of change
448from baseline in a continuous outcome measure in the clinical trial
449setting. Contemp Clin Trials 2005;26:2–16.
450[15] Dahmen G, Rochon J, Konig IR, Ziegler A. Sample size calculations for
451controlled clinical trials using generalized estimating equations (GEE).
452Methods Inf Med 2004;43:451–6.
453[16] Twisk JWR, Smidt N, de Vente W. Applied analysis of recurrent events: a
454practical overview. J Epidemiol Commun 2005;59:706–10.
455[17] Zeger SL, Liang KY . L ongitudinal data analysis for discrete and
456continuous outcomes. Biometrics 1986;42:121–30.
457[18] STATA Corporation. Intercooled STATA for Windows. Texas: College
458Station; 2007.
459[19] Capraro J, Magni C, Scarafoni A, Duranti M. Susceptibility of lupin γ-
460conglutin, the plasma glucose-lowering protein of lupin seeds, to
461proteolytic enzymes. J Agric Food Chem 2009;57:8612 –6.
462[20] Capraro J, Clemente A, Rubio LA, Magni C, Scarafoni A, Duranti M.
463Assessment of the lupin seed glucose-lowering protein intestinal
464absorption by using in vitro and ex vivo models. Food Chem 2011;125:
4651279–83.
466
Table 2t2:1
GEE analysis on human plasma glucose and insulin values adjusted by time and body mass index (BMI).
t2:2
t2:3Blood
concentrations
Dose
concentration
Regression
coefficient
Standard
error
CI 95% p-
value
t2:4Lower Upper
t2:5Glucose (mg/dL) 750 mg −8.71 1.89 −12.42 −4.99 b0.001
t2:61500 mg −11.35 3.00 −17.23 −5.46 b0.002
t2:73000 mg −8.98 2.83 −14.53 −3.43 b0.001
t2:8Insulin (μIU/mL) 750 mg 1.13 0.93 −0.69 2.94 0.224
t2:91500 mg 0.71 1.50 −2.24 3.66 0.635
t2:10 3000 mg 0.99 1.17 −1.31 3.28 0.400
6J.C. Bertoglio et al. / Fitoterapia xxx (2011) xxx–xxx
Please cite this article as: Bertoglio JC, et al., Hypoglycemic effect of lupin seed γ-conglutin in experimental animals and
healthy human subjects, Fitoterapia (2011), doi:10.1016/j.fitote.2011.05.007