Journal of Histochemistry & Cytochemistry 59(3) 328 –335
© The Author(s) 2011
Reprints and permission:
The enzyme arylsulfatase B (ARSB; N-acetylgalactos-
amine-4-sulfatase; ASB) removes 4-sulfate groups from the
non-reducing end of chondroitin-4-sulfate (C4S) or derma-
tan sulfate. ARSB has been characterized as a lysosomal
enzyme, and innate deficiency of ARSB is associated with
the lysosomal storage disease mucopolysaccharidosis
(MPS) VI or Maroteaux-Lamy syndrome.
Recent published work demonstrated that reduced ARSB
activity is associated with cystic fibrosis, in which there is
accumulation of sulfated polysaccharides, including the
sulfated glycosaminoglycans chondroitin sulfate and der-
matan sulfate (Bhattacharyya and Tobacman 2007;
Bhattacharyya et al. 2008; Bhattacharyya, Solakyildirim,
et al. 2010). Immunohistochemistry and confocal microscopy
have shown prominent cell membrane immunostaining for
ARSB in human colonic and bronchial epithelial cells and
cerebrovascular cells, as well as enzymatic activity in the
cell membrane fraction (Bhattacharyya, Solakyildirim,
et al. 2010; Bhattacharyya and Tobacman 2006, 2009). In the
metastatic human colonic carcinoma cell line T84, ARSB
activity was significantly less than in the non-transformed
NCM460 cell line. Silencing or overexpression of ARSB
affected colonic cell motility, as well as the expression of
protein and mRNA for MMP9 and the activation of RhoA
(Bhattacharyya and Tobacman 2009). Other investigators
Received for publication September 22, 2010; accepted November 30, 2010.
Joanne K. Tobacman, MD, Department of Medicine, University of Illinois
at Chicago, 840 S. Wood St. CSN 440, M/C 718, Chicago, IL 60612.
Extra-Lysosomal Localization of Arylsulfatase B in Human
Sanjiv V. Prabhu, Sumit Bhattacharyya, Grace Guzman-Hartman, Virgilia Macias,
André Kajdacsy-Balla, and Joanne K. Tobacman
Department of Pathology (SVP,GG-H,VM,AK-B) and Department of Medicine (SB,JKT), University of Illinois at Chicago, Chicago, Illinois, and Jesse
Brown VA Medical Center, Chicago, Illinois (SB,JKT).
The enzyme arylsulfatase B (N-acetylgalactosamine-4-sulfatase; ARSB; ASB) removes 4-sulfate groups from the sulfated
glycosaminoglycans (sGAG) chondroitin-4-sulfate (C4S) and dermatan sulfate (DS). Inborn deficiency of ARSB leads to the
lysosomal storage disease mucopolysaccharidosis VI, characterized by accumulation of sGAG in vital organs, disruption of
normal physiological processes, severe morbidity, and premature death. Recent published work demonstrated extra-lysosomal
localization with nuclear and cell membrane ARSB observed in bronchial and colonic epithelial cells, cerebrovascular cells,
and hepatic cells. In this report, the authors present ARSB immunostaining in a colonic microarray and show differences
in distribution, intensity, and pattern of ARSB staining among normal colon, adenomas, and adenocarcinomas. Distinctive,
intense luminal membrane staining was present in the normal epithelial cells but reduced in the malignancies and less in the
grade 3 than in the grade 1 adenocarcinomas. In the normal cores, a distinctive pattern of intense cytoplasmic positivity at
the luminal surface was followed by reduced staining deeper in the crypts. ARSB enzymatic activity was significantly greater
in normal than in malignant tissue. These study findings affirm extra-lysosomal localization of ARSB and suggest that altered
ARSB immunostaining and reduced activity may be useful indicators of malignant transformation in human colonic tissue.
(J Histochem Cytochem 59:328–335, 2011)
arylsulfatase B, colon, adenocarcinoma, membrane, lysosome, chondroitin-4-sulfate
Extra-Lysosomal Localization of Arylsulfatase B in Human Colonic Epithelium
have reported non-lysosomal arylsulfatase B and arylsulfa-
tase A in hepatic cells (Mitsunaga-Nakatsubo et al. 2009).
Reduced activity of ARSB was also shown in malignant
mammary cell lines compared to normal mammary cells
(Bhattacharyya et al. 2009). In MCF-7 cells, changes in
ARSB expression were associated with changes in mRNA
and protein expression of the proteoglycans syndecan-1 and
decorin, which have chondroitin-4-sulfate attachments
(Bhattacharyya et al. 2008). Decline in ARSB activity was
associated with increased content of C4S, attributable
to reduced hydrolysis of the 4-sulfate group on the non-
reducing end of C4S when ARSB was reduced. Because
this reaction is required for further degradation of C4S, C4S
can accumulate when ARSB activity declines (Ingmar and
Wasteson 1979). In other published work, increased cellular
content of C4S was associated with increased cellular
sequestration of interleukin-8 (IL-8) and of high molecular
weight kininogen (Bhattacharyya, Solakyildirim, et al.
2010; Bhattacharyya, Kotlo, et al. 2010). These findings
indicated that an increase in the cellular content of the
highly anionic C4S had profound effects on the processing
and release of vital cellular products.
In this report, ARSB immunostaining was evaluated in
normal, adenomatous, and malignant colonic tissues to
determine whether there were differences between the dis-
tribution and intensity of ARSB in the cytoplasm, luminal
membrane, and nuclei of colonic epithelium in malignant
versus normal colon tissue. ARSB activity was also com-
pared between malignant and normal colonic tissue
Materials and Methods
Colonic Tissue Cancer Microarray
Cancer tissue microarrays that included cores from normal
colonic tissue, adenomas, adenocarcinomas, and metastases
were obtained from the National Disease Research
Interchange (NDRI; Philadelphia, PA). Slides were arrayed
with duplicate tissue sections, and all specimens were
reviewed by NDRI pathologists and by pathologists at the
University of Illinois at Chicago (UIC). Information about
the cores included the diagnosis, the age of the patient at the
time of excision, patient gender and ethnicity, and NDRI
The microarray included 9 cores of normal colon, 4
cores from villous adenomas, and 58 cores from adenocar-
cinomas, including well, moderately, and poorly differenti-
ated. The microarray slides were hydrated through a xylene
and alcohol gradient, rinsed in distilled water. Antigen
unmasking used a 10× concentrated retrieval solution by
Dako (Carpenteria, CA), according to the manufacturer’s
instructions. Slides were then rinsed in phosphate-buffered
saline (PBS) for 5 min. For the demonstration of ARSB,
tissue sections were blocked with H2O2 blocking reagent for
10 min at room temperature. Slides were treated with a pro-
tein blocking solution for 10 min at room temperature, then
rinsed and incubated with arylsulfatase B polyclonal rabbit
antiserum (Open Biosystems, ThermoFisher Scientific,
Hunstville, AL; 1:100) for 30 min at room temperature.
Slides were rinsed, then treated with EnVision Plus labeled
polymer (DakoCytomation, Carpenteria, CA) for 30 min at
room temperature. ARSB staining was detected by DAB
Plus (DakoCytomation) for 10 min. Slides were rinsed in
distilled water, counterstained with hematoxylin, and dehy-
drated through an alcohol gradient and mounted with
Permount. Slides were scanned by ScanScope CS (Aperio
Technologies, Inc., Vista, CA), using the Aperio ImageScope
program (v10.0.35.1800) with magnification up to ×400.
Scale bars (100 µm) based on lymphocyte dimension of
7 µm were added to figures.
NDRI arrays were reviewed by three physician observ-
ers, who rated each core for percentage of cells with ARSB-
positive staining of cytoplasm, luminal membrane, and
nuclei (0–100%) and for the corresponding intensity on a
scale of 0 (none), 1 (weak), 2 (intermediate), and 3 (strong).
Differentiation of the adenocarcinomas was graded as 1, 2,
or 3. Representative images are presented in the figures.
ARSB Activity and Immunostaining in Normal
and Malignant Colon Tissue
Slides from frozen sections of adenocarcinomas and of
adjacent normal tissue and frozen tissue samples were
obtained through the Tumor Bank of the University of
Illinois at Chicago, under an established institutional
review board protocol to obtain de-identified human tissue
samples from colonic resections. Slides were stained for
ARSB using rabbit polyclonal antisera produced against a
unique C-terminal epitope (as above), rinsed in buffer, and
incubated with mouse anti-rabbit secondary antibody. After
rinsing, 3,3′-diaminobenzidine (DAB) was used for color
detection, and slides were rinsed in H2O, dehydrated, cov-
erslipped with Permount, and photographed with a Nikon
imaging system with no enhancement or modification of
ARSB activity was measured in the frozen tissue sam-
ples following an established method that used the substrate
4-methyumbilliferyl sulfate (Bhattacharyya and Tobacman
2007; Bhattacharyya et al. 2008; Bhattacharyya, Solakyildirim,
et al. 2010). Then, 20 µl of tissue homogenate and 80 µl of
assay buffer (0.05M Na acetate buffer, pH 5.6) were com-
bined with 100 µl of substrate (5 mM 4-MUS in assay buf-
fer) in wells of a microplate. The microplate was incubated
for 30 min at 37C, and the reaction was stopped by 150 µl
of stop buffer (Glycine-Carbonate buffer, pH 10.7).
Fluorescence was measured at 360 nm (excitation) and
465 nm (emission) in a fluorimeter (FluoStar Optima; BMG
Prabhu et al.
Labtech, Inc., Cary, NC) and normalized to the protein con-
tent of the tissue lysate using the BCA Protein Assay Kit
(Pierce, Rockford, IL). Measurements of at least six bio-
logical samples with technical replicates were averaged for
normal and malignant tissues and are expressed as nmol/mg
protein/hr ± standard deviation (SD).
Tests of statistical significance were performed using Instat
Software (Instat3; GraphPad, La Jolla, CA). For compari-
sons of ARSB activity in normal and malignant colonic
tissue, an unpaired, two-tailed t-test was performed with
Welch correction. One-way ANOVA with Tukey-Kramer
posttest for multiple comparisons was used for compari-
sons between the different grades of adenocarcinomas. A
p value ≤0.05 was considered statistically significant.
ARSB Activity in Normal and Malignant
ARSB activity was measured in samples of normal and
malignant colonic tissue. Mean value for activity in the
normal tissue was 108.2 ± 7.8 nmol/mg protein/hr (n=6),
compared to 55.1 ± 4.3 nmol/mg protein/hr in the malig-
nant tissue (n=8). Differences between the normal and
malignant tissue were statistically significant (p<0.0001,
unpaired t-test, two-tailed with Welch correction).
Proportions of epithelial to stromal cells appeared compa-
rable between the malignant and normal tissue, as shown in
representative sections of the tissue in which ARSB was
determined (Fig. 1A,B). ARSB immunostaining was more
intense in the normal than in the adjacent malignant colonic
tissue, including in nuclei, cytoplasm, and luminal mem-
brane, with no counterstaining.
Differences in Overall Pattern of ARSB
The differences in ARSB staining between normal and
malignant cores were quantified by ratings of the percent-
age positive staining cells of the cytoplasm, luminal mem-
brane, and nuclei (0–100%). The corresponding intensity of
staining was rated (0–3), as described in Materials and
Methods. These ratings are presented in Table 1 and indi-
cate significant differences between the normal and malig-
nant tissue in the overall percentage of positive cytoplasm
in the cells (higher in the malignant) and in the intensity of
the luminal membrane and nuclear intensity immunostain-
ing, which were higher in the normal than in the malignant
cores (p=0.05, p=0.01, p=0.001, respectively; unpaired
t-test, two-tailed). The adenocarcinomas differed signifi-
cantly in the intensity of cytoplasmic staining, which
increased from grade 1 to grade 3 (p=0.046, one-way
ANOVA with Tukey-Kramer posttest for multiple compari-
sons) (Table 2). Differences in percentage of luminal mem-
brane positivity and intensity trended down as grade
increased and neared statistical significance.
These differences were evident in the low magnification
images (Fig. 2A-D) that demonstrated marked differences
in the overall pattern of ARSB immunostaining in the nor-
mal tissue, villous adenoma, and adenocarcinoma. Grading
in the normal tissue was confounded by the distinctive pat-
tern of ARSB immunostaining (Fig. 2A). ARSB staining at
the luminal membrane and in the cytoplasm of the epithelial
cells that comprise the luminal surface was very intense.
Marked reduction in intensity of cytoplasmic staining
occurred abruptly between the differentiated cells at the
luminal surface and the transit-amplifying cells. Diminished
to absent staining persisted throughout the mid-portion and
lower region of the crypts, with the exception of a few
intensely positive cells at the base of the crypts and an
Figure 1. Arylsulfatase B (ARSB) activity and immunohistochemistry
in paired samples of normal and malignant human colonic tissue. (A)
Immunohistochemistry demonstrates prominent ARSB staining in
the membranes, nuclei, and cytoplasm of the normal tissue. (B) In
contrast, ARSB immunostaining in the tumor is markedly reduced.
ARSB immunostaining is brown; no counterstain is present. Scale bar
is 100 µm.
Extra-Lysosomal Localization of Arylsulfatase B in Human Colonic Epithelium
Table 2. Comparison of ARSB between Adenocarcinomas of
% Positive cells 84.9 (27.2) 87.9 (20.1) 91.4 (22.7)
% Positive cells 62.9 (30.5) 58.4 (36.0) 24.7 (41.6)
Intensity 1.8 (0.9)
% Positive cells 32.7 (23.5) 23.1 (22.0) 22.9 (23.2)
Intensity 1.4 (0.5)
1.7 (0.7) 2.2 (0.9)
1.5 (1.0) 0.9 (1.3)
1.3 (0.5)1.5 (0.2)
Values presented as mean ± (standard deviation) NS, not significant.
aNot all of the adenocarcinoma cores could be assigned a grade.
bp value, one-way ANOVA with Tukey-Kramer posttest for multiple
Table 1. Comparison of ARSB between Malignant and Normal
% Positive cells
% Positive cells
% Positive cells
Values presented as mean ± (standard deviation). NS, not significant.
aNot all sections were evaluable by these criteria.
bUnpaired t-test, two-tailed.
Figure 2. Overall pattern of arylsulfatase B (ARSB) staining in normal, villous adenomas, and adenocarcinomas. (A) Normal tissue
demonstrates a distinct pattern of ARSB staining, with increased intensity at the luminal surface where differentiated cells are present.
Along the mid-portion of the crypts, ARSB staining is absent, except for an occasional positively stained cell. At the base of the crypts,
rare ARSB-positive cells are present. (B) In the villous adenomas, the distinctive topography of ARSB staining seen in the normal crypt is
absent, and there is uniform epithelial staining. (C, D) These adenocarcinomas demonstrate loss of the distinctive overall pattern of ARSB
staining seen in the normal tissue, with low and high percentages of cytoplasmic staining and intensity. ARSB immunostaining is brown;
hematoxylin counterstain is blue. Scale bar is 100 µm.
Prabhu et al.
Figure 3. Luminal membrane staining of arylsulfatase B (ARSB). (A) In the villous adenomas, the luminal membrane staining for ARSB is
discontinuous in some regions, as marked by the thick arrows. In the normal colonic tissue, the luminal membrane stains intensely and
continuously for ARSB, as seen in Figure 2A. Thin arrows indicate examples of cytoplasmic inclusions. (B) In the adenocarcinomas, the
membrane staining is absent in most areas, with only residual fragments. ARSB immunostaining is brown; hematoxylin counterstain is
blue. Scale bar is 100 µm.