Distribution of the SGLT1 Na+glucose cotransporter and mRNA along the crypt-villus axis of rabbit small intestine

Vol. 181, No. 3, 1991 December 31, 1991

BIOCHEMICAL

AND BIOPHYSICAL RESEARCH COMMUNICATIONS Pages 1208-1217

Distribution of the SGLTl Na+/glucose cotransporter and mRNA along the crypt-v&s axis of rabbit small intestine Eun-Sil Hwang, Bruce A. Hirayama Department

Received

November

14,

and Ernest M. Wright *

of Physiology, UCLA School of Medicine Los Angeles, Ca. 90024-1751 1991

The expression of the Na+/glucose cotransporter (SGLTl) mRNA and protein along the crypt-viilus axis of the rabbit small intestine was examined using in sins hybridization and immunocytochemical techniques. We detected mRNA in the cells on the villus, but not in the crypts, and the mRNA abundance increased 6-fold from the base to the tip of the villus. SGLTl protein was restricted to the brush borders of mature enterocytes. We suggest that the high rate of sugar transport across the tips of the villus is due to the transcription of the SGLTl gene in mature enterocytes, the subsequent translation of SGLT mRNA, and the insertion direct of the functional SGLTl transporter into the brush border membrane of these cells lining the villus tip. @ 1991 Academic Press, Inc. Sugars are absorbed from the small intestine by the mature enterocytes lining the upper region of the villi (l-5).

Since the enterocytes are continuously renewed with a half-time of

a few days, this raises intriguing questions about the expression and regulation of transport proteins during the migration

and differentiation

of cells along the crypt-villus axis (6). One

may ask where are the genes transcribed, the mRNA translated, the proteins inserted in the plasma membranes, cDNA

probes (7). and antibodies

cotransporter

(SGLTl),

immunocytochemical SGLTl

and how are these processes regulated? (8) to the intestinal

With the recent advent of brush border

it is now possible to address these questions.

and in situ hybridization

Nat/glucose We have used

techniques to examine the distribution

mRNA and protein from crypt to villus in the rabbit small intestine.

suggest that transcription

of the

The results

is initiated as the enterocytes emerge from the crypt and increases

as the cells migrate up the villus. The mature enterocytes on the tip of the villus have the highest levels of protein in the brush border membrane and mRNA, indicating that the gene is transcribed membrane *

and SGLTl

Copyright All rights

is synthesized and inserted into the brush border

of the cells most active in sugar absorption.

To whom correspondence

0006-291X/91

protein

should

be addressed.

$1.50

8 1991 by Academic Press, Inc. of reproduction in any form reserved.

1208

Fax:

(213)206-5661.

Vol. 181, No. 3, 1991

BIOCHEMICAL

AND BIOPHYSICAL

RESEARCH COMMUNICATIONS

Tissue Prm Adult male New Zealand white rabbits (2.5 kg) were injected with a lethal dose of Nembutal. A segment of proximal jejunum was then quickly excised, flushed with ice-cold saline solution, and fixed with periodate-lysine-paraformaldehyde (9) for 3 hours at 22” and then overnight at 4°C. After washes in phosphate buffered saline containing 50 mM NH&l, tissue was embedded on dry ice with 0.C.T matrix (Tissue Tek), then cut into 10bm frozen sections which were mounted on gelatin-coated glass slides. The sections were stored at -70°C. Labeled Probe Prep. The sense and antisense RNA probes used for in situ hybridization experiments were made using T3 and T, promoters on the plasmid containing the cDNA coding for the SGLTl rabbit intestinal Na’/glucose cotranspotter (7). Template DNA was linearized and transcribed in vitro with a-[%] - UTP as the labeled nucleotide. The antisense probe has been shown to hybridize to a 2.2 kb mRNA from rabbit small intestine (10). The transcribed RNA was hydrolysed (11) to yield probes 100-150 nucleotides in kqth with a specie? activity of 4-6 x 108 cpm/ug. In Situ Hvbridization. The method described by Branks & Wilson (12) and Cox, et al. (11) was used as outlined by Wuenschell & Tobin (13). Briefly, 10Bm frozen sections were hydrated through a graded ethanol series and permeabilized with Triton X-100 and treated with proteinase K. After postfixation with 4% paraformaidehyde, sections were dehydrated and air-dried. The prehybridization mix contained 50% formamide, 0.3 M NaCI, 10 mM TrisCl pH 8.0, 1 mM EDTA, 1X Denhardt’s solution, 100 mM DTT, 0.2% SDS, 250 ug/ml Slides were incubated with salmon sperm DNA and 250 ug/ml poly A (11). prehybridization mix for 2 hours, and then probe was applied for 18 hours at 45°C. Slides were treated with RNAase, then submitted to one low stringency (2X SSC for 60 mm at room temperature) and one high stringency (0.1X SSC for 60 min at 55” C) wash. Finally, the slides were dehydrated through a graded ethanol series, then delipidated with xylene. Hybridized RNA was detected with Kodak nuclear emulsion NTB-2 applied directly to the slides, which were exposed for 7 days to one year at 4°C. The slides were then developed, stained with hematoxylin and eosin, and examined under light field and dark field conditions. Immunohistochemistry. The method used was that described by Lorenzsonn, et al. (14) except that a fluorescent detection procedure was employed. Briefly, background antigenic sites were blocked with 1% BSA. Tissue was permeabilized with 0.02% saponin at 37°C for 15 minutes , then incubated 1 hour with a 1:lO dilution of polyclonal antibody (Ab-E) raised to a 1Pamino acid polypeptide corresponding to ammo acid residues 402-420 of SGLTl (8). Controls were incubated with serum from nonimmunized hosts. Slides were then washed with 0.1% BSA, then treated with a 1:lOOO dilution of biotinylated goat-antirabbit F(ab’), fragments (Cappel) for 30 minutes. Fluorescein-conjugated avidin (Cappel; 1:lOOO dilution) was applied to the sections for 30 minutes which were then washed with 0.1% BSA. Specimens were examined using fluorescence and phase-contrast microscopy; photographs were taken with Kodak Ektachrome 400 film and automatic shutter speeds ranging between 15 and 90 seconds. RESULTS In Situ Hybridization.

Figure 1 shows low power micrographs of sections exposed to either

antisense (Fig. 1A) or sense (Fig. 1C) SGLTl

probes. A dense accumulation

of silver grams

Vol.

181, No. 3, 1991

BIOCHEMICAL

AND BIOPHYSICAL RESEARCH COMMUNICATIONS

Figure 1. In situ hybridization of Na’/glucose cotransportermRNA to [%I - labeledRNA probe in rabbit intestine; duration of exposurewas 12 months at 4°C. (A) Dark-field illumination of section treated with antisenseprobe. (B) Correspondingphasecontrast image. Note the low signalintensity in the crypt.whencomparedto mid-vilhrsandvillus tip regions.(c) Dark-field illumination of sectionwith senseprobe. (D) Correspondingphase contrastimage. Signalintensityover the length of the villas in the control (sense)specimen is that of the backgroundlabeling. (vt: villus tip; lp: lamina propria; ep: epitbelial cell; cr: crypt region). Bar represents50 pm. WaS

observed for the antisense probe over the cells lining the villi, but not the crypts. No

acctmn.tlation of grains above the background was recorded when sections were expose.d to semseprobes. Higher magnification (Fig2), showed a high signal intensity over enteroc:*es 1210

Vol.

181,

No.

3, 1991

BIOCHEMICAL

AND

BIOPHYSICAL

RESEARCH

COMMUNICATIONS

Figure 2. Epitbelial localization of in situ hybridized probe; duration of exposure was 12

months at 4°C. Dark-field illumination. (A) Antisense probe to the cotransporter. (B) Sense probe. At this higher magnification, the label is seen localized to the epithelial cells (ep), with the lamina propria (lp) virtually free of labeling over background. Bar represents 25 pm. probed with the antisense RNA

mRNA

(Fig. 2A).

To determine

whether or not a gradient of

abundance exists along the villus, we exposed sections for shorter times and

quantitated

the grain distribution.

This was achieved by taking higher magnification 1211

Vol.

181,

No.

3, 1991

BIOCHEMICAL

AND

BIOPHYSICAL

RESEARCH

COMMUNICATIONS

Figure 3. Cellular localization of in situ hybridized probe; 2-month exposure. Light-field

microscopy.(A) Antisenseprobe to the cotransporter. (B) Senseprobe. As in the longer exposures,the labeling is greatestover the epithelial cellswith a backgroundsignalin the laminapropria. (vt: villus tip; ep: epithelial cell; Ip: laminapropria). Bar represents 10 pm.

photographs of villi and counting silver grains over enterocytes at the base, mid-section, and tip of the villi (Fig. 3). The density of the grains from the antisense probe increased from the base to the tip of the villi (Fig. 4) Taking the difference between the grain densities of

Grains/pm2 0.6 m

0.4

ANTISENSE

n

SENSE

-

I BASE

.. .. .... .. .... .. *. . . .. . . .. *. .. MIDDLE

I

..*......... ... ... ....*.. . .. .* TIP

Fianre 4. Grain countsfrom in situ hybridization of SGLTl mRNA to antisenseand sense [%]-labeled

probes.

The counts over enteroqtes were obtained from photographsof

randomlyselectedareasof the villus base,rnidvillus, and villus tip regionsover more than 20 villi. Exposuretime was7 daysat 4°C. The backgroundcountsfor antisenseand sense were 0.04 f 0.01and 0.03 + O.OllrM’, respectivelyand are indicatedby the horizontal line. 1212

Vol.

BIOCHEMICAL

181, No. 3, 1991

AND BIOPHYSICAL RESEARCH COMMUNICATIONS

the antisense and sense probes as an index of abundance, the mature cells have 6 times more SGLTl

RNA than the immature

cells at the base of the vilhrs.

gradient was found for rat in a Northern

A similar mRNA

analysis of poly A+ RNA extracted from cells

isolated from the crypt to villus tip (Pajor, A M. & Wright, E.M. unpublished observation). Immune

. The polyclonal antibody specifically immunoreacts

a 70 kDa protein

in rabbit

intestinal

brush border membranes,

specifically with

but not basolateral

membranes (8). Sections treated with this antibody exhibited uniform labeling of the brush border membrane

of enterocytes lining the intestinal villi (Fig. 5). The intensity of the

labeling decreased towards the base of the villi and was undetectable

in the crypts. The

labeling of enterocyte brush borders was not observed with the secondary antibody alone, and was blocked by preincubating (lOOccg/ml).

the antibody with the nonadecapeptide

At this level of resolution,

no labeling

immunogen

was detected on the. basolateral

membrane,

within the enterocytes, or on goblet cells. A background labeling was seen in

the lamina

propria (Fig. 5) and the muscularis mucosa, but this was insensitive to the

absence or presence of peptide, and was observed in sections exposed to the secondary antibody alone. This background labeling most likely represents immunoreactivity the biotinylated

goat anti-rabbit

between

F(ab’), fragments and native rabbit immunoglobuiins.

In one experiment with a surgical biopsy of terminal

human ileum, we observed specific

decoration of the brush borders of enterocytes only at the very tip of the villi (not shown). This is consistent with the low rate of glucose absorption experimental

from the ileum in man and

animals.

DISCUSSION Enterocytes are responsible for the absorption of salt, water and nutrients from the small intestine.

They constitute 90% of the cells lining the surface of the villus, and they are

renewed every few days. The enterocytes originate from a continuously dividing population of stem cells in the crypt, and begin to differentiate

as they migrate out of the crypt to the

villus (6). As cells emerge from the crypts the enterocytes develop a brush border with a unique set of cytoskeletal and plasma membrane proteins. mid-villus,

the activities

aminopeptidase

of the brush border hydrolases, e.g. sucrase-isomaltase

and

N, are maximal, and soon thereafter the enterocytes exhibit their maximal

sugar absorptive capacities. process:

By the time the cells reach the

Sugars are absorbed by the mature enterocytes in a two stage

The first is the “active” transport across the brush border membrane 1213

by the

Vol.

181,

No.

3, 1991

BIOCHEMICAL

AND

BIOPHYSICAL

RESEARCH

COMMUNlCATlONS

Figure 5. Immunofiuorescent localization of SGLTI. (A) Section was treated with antibody to the SGLTl,

then labeled with biotinylated goat anti-rabbit F(ab)’ fragments and

fluorescein-conjugated

avidin.

Note the greater signal intensity at the villas tip. (B)

Control. Section was only treated with secondary antibodies. Strong signal in the lamina propria in both (A) and(B) is due to labeling of native rabbit plasma ceils by goat anti-rabbit F(ab)’ fragments. (bbm: brush border membrane; vt: villus tip; Ip: lamina propria).

Bar

represents 50 pm.

Na+ /glucose cotransporter; acra ISS the basolateral SGL ATl and GLUT2,

and the second is the movement

membrane

by facilitated

of the sugar out of the cell diffusion into blood. Both transporte :rs,

have been cloned, sequenced and expressed (7, 15, 16). 1214

Vol.

BIOCHEMICAL

181, No. 3, 1991

AND BIOPHYSICAL RESEARCH COMMUNICATIONS

In this study we have examined the location of SGLTl

mRNA and protein as the enterocyte

travels from crypt to villus. Using in situ hybridization

techniques we find (Figure l-4) that

SLGTl

mRNA is in cells at the base of the villus, and that the level of the mRNA increases

as cells migrate towards the tip of the villus. The Na+/glucose

cotransporter protein is only

found in the brush border of mature enterocytes towards the top of the villus (Figure 5). No significant level of SGLTl

protein was found in the cytoplasm or basolateral membrane

of enterocytes, or in the crypts. The simplest interpretation SGLTl

of these results is that the

gene is transcribed, the mRNA is translated and the protein is directly inserted into

the brush border plasma membrane

of mature enterocytes in a functional

form.

This

pattern of expression is quite distinct from that for the brush border proteins sucraseisomaltase (17) aminopeptidase

N (18) and villin (19). For the hydrolases and the structural

protein, the greatest abundance of mRNA was in cells at the crypt-villus junction

with a

decline in abundance towards the villus tip. However, the proteins are present in the brush border from the base to the tip of the villus. transcribed early in the differentiation

This suggests that these three genes are

of enterocytes at the crypt-villus junction, and there

is a low turnover rate of the gene products in the brush border over the subsequent life of the cell. Furthermore,

the results indicate that there is no single temporal/spatial

set-point

for the expression of all enterocyte genes. This is supported by recent experiments on sheep intestine where the expression of the Nat/glucose

cotransporter varied by more than 100

fold with diet but with no change in brush border marker enzyme activity or intestinal morphology (20). It will be interesting to follow the distribution

of mRNA and Na’/glucose

cotransporter protein as transport activity is varied by diet to gain further clues about the regulation of the SGLTl

gene.

Our immunocytochemical

results with the rabbit intestine are similar to those of Takata, et

al. (21) on rat intestine,

both differ from those reported by Haase, et al. (22) using

monoclonal

antibodies.

Our polyclonal antibodies immunoreact

with the cloned SGLTl

protein expressed in Sf9 cells (23) and bacteria (Hager and Wright, unpublished), immunoreactivity

in sheep brush borders correlates quantitatively

and

with the rate of sugar

transport over several orders of magnitude (20). Could the discrepancy be explained by the presence of two Nat/glucose

cotransporters in the intestine?

The evidence is indirect, but

we conclude that there is only one: 1) we have been unable, so far, to isolate a second clone from rabbit intestine (24); 2) rigorous kinetic analysis evidence suggests only one transporter in human intestinal

brush borders (25); and 3) a single missense mutation

accounts for glucose-galactose malabsorption the proteins recognized by the monoclonal

of SGLTl

in patients (26). Therefore, we suggest that antibodies in the crypts of the rat intestine are

not SGLTI. 1215

Vol.

181, No. 3, 1991

BIOCHEMICAL

AND BIOPHYSICAL RESEARCH COMMUNICATIONS

Finally, we infer from these studies that fully differentiated intestine retain the ability to transcribe the SGLTl insert the Na’/glucose

enterocytes on the rabbit small

gene, translate the mRNA, process and

cotransporter into the brush border membrane.

There are about 106

transporters in each mature enterocyte (5), and these account for the avid accumulation

of

the sugar in enterocy-tes at the tips of villi.

As would be expected for the absorption of

sugar into blood, the basolateral

glucose (GLUT2)

facilitated

is only found in the fully

mature enterocytes at the tips of the villi (27). It will be interesting to study the regulation of expression of SGLTl

and GLUT2

absorption is markedly up-regulated

in models such as the sheep intestine, where glucose by diet. (20). ACKNOWLEDGMENTS

With thanks to Drs. Bok, Coady, Ferraris, Letinsky, Stemini, Tobin and Wuenschell for valuable advice and assistance with aspects of these experiments, and Drs. LescaIe-Matys, Pajor and Turk for critical comments on the manuscript. Supported in part by N.I.H. grant AM 19567. REFERENCES

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20.

Kinter, W.B. and Wilson, T.H. (1965). J. Cell BioL a 19-39 Schneider, A.J., Kinter, W.B., and Stirling, C.E. (1966). iVew Engkd J. Meat =:305-312. Stirling, C.E., and Kinter, W.B. (1967). J Cell BioL 31:585-604. Stirling, C.E., Schneider, AJ., Wong, M-D., and Kinter, W.B. (1972). J. Chin Invest. a:43845 1. Stirling, C.E., (1967). J. Cell BioL 31:605-618 Gordon, J.I. (1989). J. CeU BioL X&l 187-1194. Hediger, M.A., Coady, M J., Ikeda, T.S., and Wright, E.M. (1987). Nahcre (London). %379-381. Hirayama, B.A, Wang, H.C., Smith, D.C., Hagenbuch, B.A, Hediger, M.A, and Wright, E.M. (1991). Am. J. PhysioL m:C296-C304. McLean, I.W., and Nakance, PK. (1974). J. Histochemdry and Cyt~hemistty. 2:1077-1083. Coady, MJ., Pajor, AM., and Wright, E.M. (1990). Am. 1. Physiol. ~C605C610. Cox, K.H., DeLeon, D.V., Angerer, LM., and Angerer, R.C. (1984). Dev. BioL U&485-502. Branks, P.L, and Wilson, M.C. (1968). MoL Brain Res. 1:1-16. Wuenschell, C.W., and Tobin, AJ. (1988). Neuron. 1:805-815. Lorenzsonn, V., Korsmo, H., and Olsen, W.A. (1987). Gartroenteroflogv. e2:98-105. Thorens, B., Sarkar, HS., Kaback, H.R., and Lodish, H.F. (1988). Cell. X:281-295. Fukumoto, H., Seino, S., Imura, H., Sieno, Y., Eddy, R.L, Fukushima, Y., Byers, M.G., Shows, T.B., and Bell, G.I. (1988). Proc. Nd Acud Sci (USA). j&5434-5438. Taber, P.G. (1990). B&hem. Biophys. Ra. Common JbQ:765-773. Nor&n, O., Dabelsteen, E., H@yer, P.E., Olsen, J., Sjijstriim, H., and Hansen, G.H. (1989). FEBS Letters =:107-112. Boller, K., Arpin, M., Pringault, E., Mangeat, P., and Reggio, H. (1988). Differentiation. Z:5 l-57. Shirazi-Beechey, S.P., Hirayama, B.A., Wang, Y., Scott, D., Smith, M.W., and Wright, E.M. (1991). J. Physiol, London. m:699-708. 1216

Vol.

21. 22. 23. 24. 25. 26. 27.

181,

No.

3, 1991

BIOCHEMICAL

AND

BIOPHYSICAL

RESEARCH

COMMUNICATIONS

Takata, K, Kasahara, T., Kasahara, M., Ezaki, O., and Hirano, H. (1991). Cell and Tisue Res. In the Press. Haase, W., Heitmann, K., Friese, W., Ollig, D., and Koepsell, H. (1990). Eumpem /. Cell BioL JJ:297-309. Smith, CD., Hirayama, B.A., and Wright, E.M. (1991). Biochim. Biophysu Acfa In the Press. Ideka, T.S., Hwang, E-S., Coady, M.J., Hirayama, B.A., Hediger, MA., and Wright, E.M. (1989). J. Memb. BioL JJQ:87-95. Maio, C., and Berteioot, A. (1991). /. Memb. BioL m:127-141. Turk, E., Zabel, B., Mundlos, S., Dyer, J., and Wright, E.M. (1991). Nature, London. &X&354-356. Thorens, B., Cheng, Z-Q., Brown, D., and Lotish, H.F. (1990). Am L PhysioL ~:c279-c285.

1217