Intestinal-enriched Krüppel-like Factor (Krüppel-like Factor 5) Is a Positive Regulator of Cellular Proliferation*

Intestinal-enriched Krüppel-like factor (IKLF or KLF5) belongs to the family of mammalian Krüppel-like transcription factors. Previous studies indicate that expression of IKLF is enriched in the proliferating crypt epithelial cells of the intestinal tract. However, the biological function of IKLF is unknown. In the current study, we have shown that the level of IKLF mRNA was nearly undetectable in serum-deprived NIH3T3 fibroblasts but became acutely and significantly increased upon the addition of fetal bovine serum or the phorbol ester, PMA. This induction required protein synthesis because it was prevented by cycloheximide. Transfection of IKLF into NIH3T3 cells resulted in the formation of foci in a manner similar to that caused by the activated Ha-ras oncogene. Constitutive expression of IKLF in transfected NIH3T3 cells significantly increased the rate of proliferation when compared with cells transfected with an empty vector. The growth ofIKLF-transfected cells was no longer inhibited by cell-cell contact or by low serum content. Moreover, these cells proliferated in an anchorage-independent fashion. We conclude thatIKLF encodes a delayed early response gene product that positively regulates cellular proliferation and may give rise to a transformed phenotype when overexpressed.

ily of Krü ppel-like factors (KLFs) 1 that are highly related to the mammalian Krü ppel protein erythroid Krü ppel-like factor (EKLF) (4) have been described (2,5,6). Many of these proteins were given a numerical designation by the Human Gene Nomenclature Committee (HGNC, Ref. 7), with EKLF designated as KLF1. The genes encoding many KLFs are expressed in a tissue-specific or -selective manner. In addition, evidence suggests that KLFs collectively exert important regulatory functions in diverse biological processes such as growth, development, differentiation, and apoptosis.
One tissue in which a number of KLFs appear to play an important regulatory role is the intestinal epithelium. This tissue is a dynamic system in which proliferation of stem cells located in the crypts is intimately coupled to their differentiation into mature daughter cells once they exit the crypts (8,9). Expression of the genes encoding two KLFs, gut-enriched Krü ppel-like factor (GKLF or KLF4, Ref. 10,11) and intestinal-enriched Krü ppel-like factor (IKLF or KLF5, Refs. 12,13), is particularly active in the intestinal epithelium. However, their patterns of expression appear to be complementary rather than redundant. Whereas GKLF is primarily expressed in the differentiated epithelial cells, away from the proliferating zone, IKLF is found mainly in the proliferating crypt cell population. The in vivo pattern of GKLF expression is mirrored by its in vitro pattern; it is found mostly in cells that are growtharrested (10,14). Moreover, constitutive expression of GKLF leads to the inhibition of DNA synthesis (10,15). In contrast, the physiological function of IKLF is less clear, although it has been proposed to have an opposing effect to GKLF in regulating epithelial cell differentiation (12).
The present study seeks to characterize the effect of IKLF on cellular proliferation. We demonstrate that expression of IKLF in cultured cells responds transiently and acutely to growth stimuli. In addition, forced expression of IKLF in transfected cells results in an accelerated rate of proliferation and a transformed phenotype as evidenced by formation of foci, loss of contact inhibition, as well as acquisition of serum-and anchorage-independent growth. Our results indicate that IKLF has a proproliferative effect, which lends support to the previous hypothesis that it may counteract the function of GKLF.
Cell Culture-NIH3T3 cells were maintained in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% FBS, 100 g/ml streptomycin, and 100 units/ml penicillin at 37°C in a 5% CO 2 atmosphere. For experiments involving mitogenic stimulation, cells were first rendered quiescent by removal of serum from the medium for 24 h. Cells were then stimulated with fresh medium containing 15% FBS or 100 ng/ml PMA for various duration. In experiments in which CHX was included, quiescent cells were pretreated with 10 g/ml CHX for 1 h before the addition of FBS or PMA. Treatments were then continued for another 1 h with the respective mitogen and CHX.
Northern and Western Blot Analyses-RNA was isolated using the Trizol method (Life Technologies) and resolved by denaturing agarose * This work was supported in part by Grants DK52230 and CA84197 from the National Institutes of Health and the John G. Rangos Charitable Foundation. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Focus Formation Assay-Focus formation assays were performed according to a previously published protocol (17). Briefly, 5 g of plasmid DNA containing the various effectors was transfected into NIH3T3 cells using LipofectAMINE (Life Technologies, Inc.). Cells were then maintained in DMEM and 10% FBS for 2-3 weeks at which time they were stained with methylene blue to demonstrate the foci. The number of foci in each dish was then manually counted.
Establishment of Stable IKLF-expressing Cell Lines-NIH3T3 cells were transfected with pBK-CMV-IKLF⅐HA or the pBK-CMV empty vector using LipofectAMINE. Two mg/ml G418 was added to the medium beginning 24 h after transfection to select for resistant clones, which were isolated 2 weeks later and expanded. The presence of HA-tagged IKLF was detected by Western blot analysis using the anti-HA monoclonal antibody.
Cell Proliferation and Serum and Anchorage Dependence Assays-Cell proliferation assays were performed by seeding the IKLF⅐HA-or empty vector-transfected cells at a density of 2 ϫ 10 5 cells per well in 6-well plates. Cells were fed DMEM with 10% FBS every other day, and the number of cells in the wells were manually counted with a hemocytometer daily for up to 5 days following seeding. For the serum dependence assay, cells were seeded at a density 2 ϫ 10 5 cells per well in 6-well plates and maintained in DMEM with 1% FBS. Cells were counted every other day up to 6 days after seeding. Anchorage dependence assay was performed according to a published protocol (17). Haras-, IKLF⅐HA-, or empty vector-transfected cells were seeded at a density of 5 ϫ 10 5 cells per 10-cm dish in a 0.33% top agar suspension, which was overlaid onto a 0.5% agar bottom layer. Cells were fed DMEM, 10% FBS, and 2 mg/ml G418 every other day. Colonies developed in the agar suspension were examined 3 weeks following seeding, and the number was tabulated under an inverted phase-contrast microscope. Photodocumentation was accomplished with a Nikon digital camera.

IKLF Encodes a Delayed Early Response Gene to Growth
Stimulation-To investigate the responsiveness of IKLF expression to growth stimulation, we conducted Northern blot analyses in cultured fibroblasts. NIH3T3 cells were first rendered quiescent by removing the serum from the medium and then stimulated by 15% FBS or the phorbol ester, PMA. As seen in Fig. 1, A and B, the level of IKLF transcripts was barely detectable in serum-starved, quiescent cells (time 0). Upon addition of FBS (Fig. 1A) or PMA (Fig. 1B), the levels of IKLF transcripts increased acutely and transiently, reaching a maximum after 2-3 h of treatment before returning to baseline levels. The enhancing effect of both stimuli was prevented in cells pretreated with the protein synthesis inhibitor, cycloheximide (Fig. 1C, CHX). These results indicate that IKLF is an early response gene to growth stimulation although this responsiveness requires protein synthesis. IKLF would therefore fall into the category of delayed early response genes (18,19).
Forced Expression of IKLF Causes Formation of Foci-As a means to measure the effect of IKLF on cell proliferation, we conducted the focus formation assay previously developed for testing the transforming activity of oncogenes such as ras (17,20). NIH3T3 fibroblasts were transfected with the various effector constructs and foci scored 2-3 weeks after transfection. As shown in Fig. 2, an expression plasmid containing the HAtagged IKLF produced ϳ30% of the number of foci caused by activated Ha-Ras. A second construct containing IKLF without the HA-tag gave rise to similar results (not shown). In contrast, an expression plasmid containing GKLF failed to produce any foci, as were mock-transfected cells (Fig. 2). These results suggest that forced expression of IKLF causes focus formation in a manner similar to activated Ha-Ras, an activity that was not paralleled by GKLF.
Constitutive Expression of IKLF Causes Accelerated Cell Growth-To further investigate the effect of IKLF on cell proliferation, we established several clonally derived NIH3T3 cell lines that had been transfected with pBK-CMV-IKLF⅐HA or the pBK-CMV empty vector and selected with the antibiotic, G418. Two independent clones from each construct were chosen FIG. 1. Activation of IKLF expression in NIH3T3 cells upon mitogenic stimulation. Before addition of mitogens, cells were incubated in medium stripped of serum for 24 h. Medium containing fresh FBS (15% v/v, A) or PMA (100 ng/ml, B) were then added to the cells for the various times indicated. RNA was extracted from the cells and probed simultaneously for the IKLF and GAPDH transcript content. In C, serum-deprived cells were preincubated in the absence (lanes 1, 3, and 5) or presence (lanes 2, 4, and 6) of 10 g/ml CHX for 1 h, followed by an additional 1 h of medium alone (lanes 1 and 2), medium containing FBS (lanes 3 and 4), or PMA (lanes 5 and 6) in the absence or presence of CHX. RNA was extracted and probed for both IKLF and GAPDH.

FIG. 2. Focus formation assay.
Focus formation assay was conducted as described under "Experimental Procedures." NIH3T3 cells were transfected with 5 g/10-cm dish of an expression plasmid containing Ha-Ras (16), IKLF⅐HA (12), or GKLF (10). The control was mock-transfected NIH3T3 cells. Foci were visualized by staining with methylene blue 2-3 weeks following transfection. Shown are the means of four independent experiments, each conducted in quadruplicate. Bars indicate S.E. *, p Ͻ 0.01 when compared with control, mocktransfected cells. and examined. Fig. 3A shows that the two pBK-CMV-IKLF⅐HA-but not the two pBK-CMV-transfected clones ( lanes  1 and 2 versus lanes 3 and 4, respectively) produced a fulllength HA-tagged IKLF as detected by Western blot using a monoclonal antibody against the HA epitope. When the growth characteristics of these cells were examined and compared over a course of 5 days following seeding at a low density, it became apparent that the IKLF⅐HA-expressing cells proliferated at a much faster rate compared with the control, empty vectortransfected cells (Fig. 3B). Moreover, whereas the control cells ceased to proliferate after reaching confluency, the IKLF⅐HAtransfected cells continued to grow to several layers (Fig. 3C). These findings suggest that the growth of IKLF⅐HA-transfected cells was no longer subject to contact inhibition, providing further evidence for a proproliferative effect of IKLF.
IKLF Causes Serum-and Anchorage-independent Growth-The stably transfected NIH3T3 cells were subjected to additional and more stringent tests of cellular proliferation and transformation. In one, cells were seeded and maintained in medium containing only 1% FBS. Under these conditions, empty vector-transfected cells failed to proliferate. In fact, many perished because of the low-serum content. In contrast, IKLF⅐HA-transfected cells continued to proliferate at a relatively brisk rate albeit slightly slower than that when maintained in 10% FBS (compare Figs. 4A and 3B). A second test involved growth in soft agar (17). As seen in Fig. 4B, IKLF⅐HAtransfected cells formed colonies in an agar suspension as were Ha-ras-transfected cells at a ϳ1:3 ratio. In contrast, empty vector-transfected cells remained as single cells in the agar suspension without ever forming any colonies (Fig. 4C). The morphology and size of the colonies produced by IKLF⅐HA-and Ha-ras-transfected cells were very similar to each other (Fig.  4C).

COMPLEMENTARY DISCUSSION
Complementary DNA clones encoding mouse IKLF were initially identified because of sequence homology to LKLF (12). A human homolog was subsequently isolated based on its binding to a specific cis-sequence in the promoter of the lactoferrin gene promoter (21). In situ hybridization studies of both adult (12) and fetal intestinal tissues (13) showed that expression of IKLF is concentrated in the base of intestinal crypts. The in vivo pattern of IKLF expression in the intestinal tract therefore correlates with a proliferative phenotype, although a direct effect of IKLF on cellular proliferation was not established by these studies. It should be noted that IKLF is identical to the previously isolated BTEB2 (22), which, because of a sequencing error, has a shorter open-reading frame than IKLF (13). In a rabbit model, expression of BTEB2/IKLF is induced in activated smooth muscle cells (SMCs) in the neointima of balloon- injured aorta (23). Similarly, increased BTEB2 expression has been noted in proliferating SMCs at anastomotic vascular stricture (24) and this increased expression is a positive predictive factor for vascular restenosis in pathological conditions (25). Taken together, these studies suggest that activation of BTEB2/IKLF expression is correlated with a proliferative state.
The results of the current study indicate that expression of IKLF in cultured, quiescent NIH3T3 cells is acutely and transiently induced upon mitogenic stimulation by factors such as serum and phorbol ester. This induction is dependent on new protein synthesis as it is abolished in the presence of cycloheximide. A similar inductive response of BTEB2/IKLF was noted in cultured rabbit aorta-derived SMCs treated with PMA or basic fibroblast growth factor (26). The latter study also showed that expression of the immediate early response gene, Egr-1, is highly up-regulated by PMA and that Egr-1 binds to and activates the promoter of the BTEB2/IKLF gene (26). It is therefore possible that Egr-1 is an immediate mediator of induction of IKLF during proliferative responses.
Despite evidence from in vitro and in vivo studies demonstrating a correlation between IKLF expression and proliferation, it is not clear whether IKLF directly regulates cellular proliferation. The results of the current study are the first to show that constitutive expression of IKLF increases proliferation. We also show that IKLF alone is sufficient to cause a transformed phenotype as assessed by focus formation (Fig. 2), loss of contact inhibition (Fig. 3C), and gain of serum- (Fig. 4A) and anchorage-independent growth (Fig. 4, B and C). These observations therefore suggest that IKLF is potentially a mediator of cellular proliferation in the various in vivo and in vitro systems described above. Whether IKLF directly participates in regulating the cell cycle or whether expression of IKLF is increased in pathological conditions such as neoplasm remains to be investigated.
Based on the opposing patterns of expression of GKLF and IKLF in the intestine, Lingrel and co-workers (12) proposed that their gene products may have opposing effects in regulating proliferation and differentiation of the intestinal epithelium. Indeed, the antiproliferative effect of GKLF depicted by previous studies (10,14) and the proproliferative effect of IKLF demonstrated by this study support their hypothesis. Biochemical evidence also suggests that the two proteins may have opposing functions. For example, the promoter of the SMC differentiation marker gene, SM22␣, is repressed by GKLF but activated by IKLF (27). Preliminary studies from our laboratory also indicate that GKLF and IKLF regulate the GKLF promoter in an opposing manner. 2 In view of the highly conserved sequence in the zinc finger region of the two proteins and the similar DNA sequences to which they bind (10,11,12,14,21), it would not be surprising that the two KLFs may coordinately regulate the expression of a group of genes through similar if not identical cis-elements. Further studies will demonstrate the biochemical mechanisms by which GKLF and IKLF antagonize each other in the context of regulating complex biological processes such as cellular proliferation and differentiation.