Mucin-depleted foci (MDF) are microscopic dysplastic lesions induced in the colon of rodents by specific colon carcinogens. Most MDF show Wnt pathway activation, whereas only a subset shows mutations in the Ctnnb1 gene, coding for β-catenin. Because Apc is a member of the Wnt pathway and the most frequent mutated gene in human colon cancer, we tested whether MDF harbor Apc mutations. F344 rats were treated twice with 150 mg/kg of 1,2-dimethylhydrazine. After 15 or 28 weeks, MDF, aberrant crypt foci (ACF), and tumors were collected. We screened a segment of the Apc gene comprising the region homologous to the mutation cluster region (MCR) of human APC, which frequently shows mutations in experimental colon tumors. Mutations were identified by PCR amplification and sequencing in 6:24 MDF (25%), 7:23 tumors (30%), 0:24 ACF (0%). Most of the mutations (92%) in MDF and tumors were localized in a region upstream from the MCR. All mutations were single-base substitutions and mainly formed by G:C → A:T and C:G → T:A transitions. The pattern of nucleotide changes was similar in MDF and tumors, and, interestingly, the same mutation in codon 1047 was found in two MDF and in three tumors. Four out of the six mutations found in MDF were nonsense mutations, and two were missense. All mutations in tumors determined a protein truncation. These results show that Apc mutations are present in MDF with a frequency similar to that of tumors, strengthening the evidence that they are precancerous lesions in colon carcinogenesis. [Cancer Res 2007;67(2):445–9]

Mucin-depleted foci (MDF) are dysplastic lesions originally identified in the colon of azoxymethane-treated rats (1). MDF are dose-dependently induced by colon carcinogens (2), and their occurrence is correlated to the number of tumors after the administration of promoters or chemopreventive agents (36), thus suggesting that these lesions are preneoplastic. We previously reported that most MDF, like tumors, show constitutive activation of the Wnt-signaling pathway with cytoplasmic and nuclear accumulation of β-catenin (2). However, because only a fraction of MDF (25%) harbors mutations in the rat gene coding for β-catenin (Ctnnb1 gene), other components of the Wnt pathway might be altered in these lesions.

The Apc gene is a component of the Wnt-signaling pathway as a member of the macromolecular complex which degrades β-catenin (7). Mutations in either the Apc or Ctnnb1 gene make β-catenin resistant to degradation, thus causing its accumulation in the nuclei and activation of the Wnt-signaling pathway. Germ line APC mutations are responsible for the inherited predisposition for familial adenomatous polyposis (FAP) and for intestinal tumors in the Min mouse carcinogenesis model (8, 9). Somatic APC mutations are also common in sporadic colorectal cancers (8) and are present with variable frequencies in chemically induced colon cancer of rodents (10, 11).

On the basis of these considerations, we studied Apc mutations in rat MDF as a possible cause of Wnt pathway activation observed in these lesions. Moreover, the presence of a mutation so relevant for colon carcinogenesis such as the Apc would strengthen the hypothesis that MDF are preneoplastic. To compare MDF with more advanced stages of transformation or with other purported preneoplastic lesions, we also studied Apc mutations in colon tumors and aberrant crypt foci (ACF), well characterized in rodents and humans (12, 13).

Carcinogenesis induction. Male F344 rats, 4 to 5 weeks old, were obtained from Nossan (Correzzana, Milan, Italy). They were housed according to the European Union Regulations on the Care and Use of Laboratory Animals (14) and fed a high-fat diet (23% corn oil w/w) based on the American Institute of Nutrition-76 diet (2). The experimental protocol was approved by the Commission for Animal Experimentation of the Italian Ministry of Health. At 6 to 7 weeks of age, rats were treated twice, 1 week apart, with s.c. injections of 1,2-dimethylhydrazine (DMH; 150 mg/kg × 2). Rats were divided into two groups. One group was sacrificed by CO2 asphyxiation 15 weeks after the first injection, and the other was sacrificed 28 weeks afterward. The colons were excised and opened longitudinally. Those lesions which were visible by the naked eye at sacrifice were defined as tumors following subsequent histopathologic examination as described (2). To collect MDF, ACF, or tumors, the colons were then fixed flat, as reported (1, 2), using 70% ethanol for no more than 3 h.

Identification of ACF and MDF. MDF are foci of crypts depleted of mucins in a background of normal crypts using the high-iron diamine Alcian blue (HID-AB) technique, which allows differentiation between sialomucin and sulfomucin production (1). MDF have also been identified by other authors using Alcian blue (AB), which stains in blue all acid mucins (6). However, we found that both staining methods (HID-AB or AB) degrade DNA, making it unsuitable for relatively long amplification by PCR (data not shown). Therefore, we used a milder staining procedure (Alcian blue-neutral red [AB-NR]) consisting of a shorter AB incubation (10 min versus 30 min for standard AB staining), followed by a brief counterstaining with neutral red (0.5% aqueous solution for 10 s) to identify MDF in this study. In fact, with the AB-NR staining, the epithelial layer of the normal mucosa is shown as a reddish background dotted with blue spots representing the opening of normal crypts full of mucus stained with AB. In this blue-dotted background, MDF is shown as a reddish spot in which the crypts, often distorted and smaller than normal, do not produce mucins (therefore devoid of blue; Fig. 1A). The correspondence between the original identification method (HID-AB) and the AB-NR method was studied in seven rat colons in which all MDF observed with the AB-NR staining were confirmed as MDF with a subsequent HID-AB staining (62 out of 62 total MDF scored in AB-NR). An example of this correspondence is shown in Fig. 1 (A and B). ACF were also clearly visible using AB-NR as foci elevated from the plane of the mucosa, formed by crypts larger than normal, with increased pericryptal space and a thicker layer of epithelial cells (Fig. 1C). Each lesion identified as MDF or ACF in AB-NR–stained colons was marked with permanent ink and processed as previously described (2).

Figure 1.

Microscopic appearance of MDF or ACF in Alcian blue-neutral red (AB-NR)– or HID-AB–stained unsectioned colons. A, identification of an MDF (yellow arrow) in Alcian blue-neutral red–stained colon. B, the same MDF (yellow arrow) observed in A after restaining of the colon with the HID-AB technique. C, typical appearance of an ACF (yellow arrow) in AB-NR–stained colon. Original magnification, ×100.

Figure 1.

Microscopic appearance of MDF or ACF in Alcian blue-neutral red (AB-NR)– or HID-AB–stained unsectioned colons. A, identification of an MDF (yellow arrow) in Alcian blue-neutral red–stained colon. B, the same MDF (yellow arrow) observed in A after restaining of the colon with the HID-AB technique. C, typical appearance of an ACF (yellow arrow) in AB-NR–stained colon. Original magnification, ×100.

Close modal

Apc analysis. ACF, MDF, and tumors were laser microdissected (PALM MicroBeam system) from longitudinal paraffin 5-μm-thick sections (two to four sections per lesion), and the DNA was extracted using the Qiamp DNA Micro Kit (Qiagen, Chatsworth, CA) according to the manufacturer's instructions. We analyzed a fragment of the rat Apc gene encompassing the homologous region of the human MCR (ref. 8; nucleotides [nt] 3,900–4,589 of the rat sequence, GenBank accession no. D38629) and a region upstream of the MCR (nt 3,078–3,899). Importantly, this segment comprises the region from nucleotides 3,078 to 3,997, where the majority of Apc mutations have been described in experimentally induced rodent tumors (10, 11). The Apc gene region was divided in four overlapping fragments (fragments 1–4) for mutational analysis. Due to the very low amount of DNA that could be extracted from microdissected MDF and ACF, a nested PCR approach was planned. For each lesion, in the first round of amplification, the entire region of the Apc gene was obtained by two duplex PCR in which two nonoverlapping fragments were coamplified in the same reaction. In the second round of PCR amplification, each fragment was amplified separately by nested primers using a dilution of the first PCR amplification as template. The primer sequences were designed using the published rat Apc cDNA sequence (GenBank accession no. D38629). The primers used for the first round of amplification were fragment 1 (nt 2,900–3,507): 1Fw: 5′-TGCCAAGGTGGAATATAAGAGA; and 2Rv: 5′-CATGCTGTTCTTCCTCAGAATA; fragment 2 (nt 3,260–3,859): 3Fw: 5′-GAACACCAATTTTCCCGTCT; and 4Rv: 5′-GAGAAACATATTGGGGTGTCTT; fragment 3 (nt 3,696–4,226): 5Fw: 5′-GCTGCATCTGCACCTTCAT; and 6Rv: 5′-TGGAGCGACTCTCAAAACTG; and fragment 4 (nt 4,033–4,662): 7Fw: 5′-CTAGAACCAAGCCCAGCAGA; and 8Rv: 5′-CAGTTTCATTCCCATTGTCG.

The primers used in the nested amplifications were fragment 1Nested (nt 2,985–3,455): 1NFw: 5′-AGCCCTCGGTTGAGTCCTAC; and 2NRv: 5′-TTCATAGTCGTCTTCCTGACACA; fragment 2Nested (nt 3,322–3,811): 3NFw: 5′-ATTTTGGGCAACAAGAGTGC; and 4NRv: 5′-TGGTTGATGGAGGGGACTT; fragment 3Nested (nt 3,768–4,197): 5NFw: 5′-CAGACTCCAAAAGGGACTGC; and 6NRv: 5′-TGAATACAAGGGGCGTCTCC; and fragment 4N (nt 4,091–4,640): 7NFw: 5′-TAAAGCTGTTGAGTTTTCTTCAG; and 8NRv: 5′-TTCCTGAACTGGAGGCATTA.

PCR reactions were carried out in a 25-μL volume containing 1× PCR buffer, 1.5 mmol/L MgCl2, 0.3 mmol/L deoxynucleotide triphosphates, 0.4 μmol/L of each primer and 1.25 units of Taq polymerase (Advanced Biotechnologies, Epsom, United Kingdom). Cycling conditions were initial denaturation for 5 min at 94°C, 35 cycles of denaturation for 1 min at 95°C, annealing for 1 min at 58°C, and extension at 72°C for 1 min with a final extension step for 10 min at 72°C. The final products were separated on a 2.5% agarose gel and visualized by ethidium bromide staining. Final PCR products from nested amplifications were purified using the Wizard SV Gel and PCR Clean-Up System (Promega, Madison, WI) and directly sequenced in both directions using the ABI Prism Big Dye Terminator Cycle Sequencing Ready Reaction Kit (Applied Biosystems, Foster City, CA); sequences were analyzed using the DNA sequencer ABI Prism 310 Genetic Analyzer (Applied Biosystems) following the manufacturer's protocol.

Statistical evaluation of the data.Apc mutation frequencies in ACF, MDF, and tumors were evaluated with Fisher's exact test, followed by the determination of odds ratios (OR) to evaluate the statistical significance of the difference between lesions. A trend analysis (Cochran-Armitage trend test) was also used to evaluate the frequencies of mutations in the three different lesions (ACF, MDF, and tumors). All the statistical calculations were done using Exact Methods Statxact6 statistical package (Cytel Software Corporation, Cambridge, MA).

Apc mutational analysis in tumors. Twenty-three tumors (adenomas n = 12; and adenocarcinomas n = 11) were analyzed for Apc mutations. Seven of these tumors were mutated (frequency of mutated lesions, 30.4%); four out of the seven Apc mutated tumors were adenomas, and three were adenocarcinomas (Table 1). All mutations were single-base changes resulting in a stop codon. Six mutations were transitions (three G:C → A:T and three C:G → T:A), one was a G:C → T:A transversion (Table 1). All Apc mutations were found in a region between nt 3,186 and 3,810, upstream of the homologous region of the human MCR (nt 3,900–4,589); (Table 1; Fig. 2).

Table 1.

Apc mutations in MDF and tumors induced by DMH in rat

LesionHistologyCrypts/focusNucleotide positionAmino acid positionNucleotide changeAmino acid change
Tumors Adenocarcinoma  3,186 1,045 GAA → TAA Glu → Stop 
 Adenoma  3,194 1,047 TGG → TGA Trp → Stop 
 Adenoma  3,194 1,047 TGG → TGA Trp → Stop 
 Adenocarcinoma  3,194 1,047 TGG → TGA Trp → Stop 
 Adenoma  3,246 1,065 CAA → TAA Gln → Stop 
 Adenocarcinoma  3,387 1,112 CGA → TGA Arg → Stop 
 Adenoma  3,810 1,253 CAA → TAA Gln → Stop 
MDF  80 3,084 1,011 CAT → TAT His → Tyr 
  15 3,194 1,047 TGG → TGA Trp → Stop 
  3,194 1,047 TGG → TGA Trp → Stop 
  11 3,231 1,060 CAA → TAA Gln → Stop 
  3,822 1,257 CAG → TAG Gln → Stop 
  4,444 1,464 GGT → GAT Gly → Asp 
LesionHistologyCrypts/focusNucleotide positionAmino acid positionNucleotide changeAmino acid change
Tumors Adenocarcinoma  3,186 1,045 GAA → TAA Glu → Stop 
 Adenoma  3,194 1,047 TGG → TGA Trp → Stop 
 Adenoma  3,194 1,047 TGG → TGA Trp → Stop 
 Adenocarcinoma  3,194 1,047 TGG → TGA Trp → Stop 
 Adenoma  3,246 1,065 CAA → TAA Gln → Stop 
 Adenocarcinoma  3,387 1,112 CGA → TGA Arg → Stop 
 Adenoma  3,810 1,253 CAA → TAA Gln → Stop 
MDF  80 3,084 1,011 CAT → TAT His → Tyr 
  15 3,194 1,047 TGG → TGA Trp → Stop 
  3,194 1,047 TGG → TGA Trp → Stop 
  11 3,231 1,060 CAA → TAA Gln → Stop 
  3,822 1,257 CAG → TAG Gln → Stop 
  4,444 1,464 GGT → GAT Gly → Asp 
Figure 2.

Schematic representation of the Apc region analyzed and localization of the mutations found in MDF and tumors. The nucleotide numbers are assigned according to the published rat sequence (GenBank accession no. D38629). *, missense mutations.

Figure 2.

Schematic representation of the Apc region analyzed and localization of the mutations found in MDF and tumors. The nucleotide numbers are assigned according to the published rat sequence (GenBank accession no. D38629). *, missense mutations.

Close modal

Apc mutational analysis in MDF and ACF. Six out of the 24 MDF studied harbored Apc mutations (frequency of mutated lesions, 25%). All mutations were single-base substitutions resulting in transition mutations (three G:C → A:T and three C:G → T:A) (Table 1). Four of these mutations resulted in a stop codon, whereas two mutations were missense (Table 1). All but one mutation were localized upstream of the homologous region of the human MCR (Table 1; Fig. 2).

Most (four out of six) of the mutated MDF were formed by <15 crypts (11, 7, 6, and 9 crypts/focus); only one had a multiplicity of about 80 crypts/focus [the mean number of crypts forming each MDF: 15.50 ± 4.8 (SE); range, 3–100].

The same mutational analysis of the Apc gene was also done in 24 ACF (mean multiplicity, 8.0 ± 0.5 SE; range, 4–13). None of the ACF analyzed had Apc mutations.

The mutation frequencies of the different lesions (0:24, 6:24, and 7:23 in ACF, MDF, and tumors, respectively) were significantly different (P = 0.006 by Fisher's exact test); differences between ACF and MDF, ACF and tumors were statistically significant (exact P values for OR were 0.02 and 0.008, respectively), whereas the frequency of mutation in MDF and tumor was similar (exact P value for OR, 0.93). Moreover, a trend analysis calculated on the frequencies of mutation in the three different lesions (ACF, MDF, and tumors) showed that the trend was statistically significant (exact P = 0.011 for Cochran-Armitage trend test).

The aim of this study was to test whether MDF were mutated in Apc, a key gene in colorectal carcinogenesis. We analyzed a region homologous of the human MCR and a segment upstream, where the majority of mutations in carcinogen-induced rats have been reported (10, 11). About 30% of all the tumors we analyzed (adenomas and adenocarcinomas) harbor mutations in the Apc gene. The mutations were nonsense and were located upstream (nt 3,186–3,810) of the homologous region of the human MCR (nt 3,900–4,589), a finding in agreement with previous experimental studies (10, 11). Interestingly, all but one mutation were found in the region (nt 3,150–3,452) between the first two human homologous 15-amino acid repeats for β-catenin binding (Fig. 2), thus suggesting that mutations in this location may induce a growth advantage in our experimental model.

Our results also clearly show that 25% of the MDF analyzed harbor Apc mutations, a frequency similar to that found in tumors. As observed in tumors, the mutations in MDF were mainly located upstream of the MCR (Table 1; Fig. 2) and had the same nucleotide change profile as tumors. In fact, 42% of tumor mutations were G:C → A:T and 42% C:G → T:A. Similarly, MDF mutations were 50% G:C → A:T and 50% C:G → T:A. Interestingly, two MDF harbored the same mutation (TGG → TGA, codon 1,047) found in three tumors (two adenomas and one adenocarcinoma), thus suggesting that at least some MDF may be true tumor precursors.

We also found that not all MDF mutations resulted in a truncated protein because two out of the six mutations found were missense (Table 1). One of the two missense mutations was found in a rather large MDF formed by about 80 crypts, whereas the other one was carried by an MDF formed by 7 crypts. Although it has been reported that most mutations in the Apc gene result in a truncated protein (15), it is worth noting that a previous experimental study reported a high frequency of missense mutations in DMH-induced colon tumors (10), whereas a recent study of human colorectal tumors (16) reported the same.

Regarding ACF, none of the lesions tested had Apc mutations, a result in agreement with a previous study by our group in which these lesions were analyzed using the protein truncation test (11). ACF are putative preneoplastic lesions that have been extensively characterized in rodents and humans (12, 13). Although ACF show preneoplastic characteristics, they are heterogeneous lesions that often regress to a more normal phenotype, whereas only a few progress to more advanced stages. We previously reported that ACF show less dysplasia than MDF (1), a phenomenon that could be related to the absence of Apc mutation. In fact, Apc mutations in ACF have been reported in human studies in very large dysplastic ACF (about 200 crypts; ref. 17) or in one dysplastic ACF formed by about 30 crypts in another study (18). However, other authors have reported lack of Apc mutations in ACF irrespective of their dysplasia (19).

Our data do not permit to evaluate how tumors develop from preneoplastic lesions; however, it is a fact that only a fraction of all the lesions formed after carcinogen administration evolves into tumors. Previous results from our group (2) showed that in DMH-treated rats, the number of MDF increases after carcinogen treatment, but decrease later, when macroscopic tumors appear; this suggests that at least some MDF may transform into tumors (2). The results obtained in the present study suggest that the MDF tumor transformation may be primed by Apc mutations, occurring quite early because many mutated MDF were formed by <15 crypts. As for the relation between ACF and MDF, we found a significantly different Apc mutation frequency between these two lesions. Although the analysis of the data showed a significant trend of Apc mutation in the sequence “ACF-MDF tumors,” we cannot say at the moment whether MDF evolve from ACF acquiring mutations in Apc or in other cancer-related genes, or if they develop independently from ACF.

In conclusion, the fact that Apc mutations occur in MDF and tumors with a similar frequency indicates that these lesions are closely related. Moreover, these results, together with our previous findings (2), indicate that Ctnnb1 (β-catenin) and Apc mutations in MDF are the main determinants for Wnt pathway activation, a phenomenon that can also induce a down-regulation of mucin production (20). Thus, the loss of a physical and chemical barrier of mucus could lead MDF in strict contact with inflammatory stimuli or food mutagens and increase the probability of acquiring further mutations and a more advanced transformed phenotype.

Grant support: American Institute for Cancer Research (grant 05A019-REV), by Associazione Italiana per la Ricerca sul Cancro (Italian Association for Cancer Research, Regional Grant) and by Fondo Ateneo ex-60% of the University of Florence.

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.

We thank M. Beni and P. Ceccatelli for their expert technical assistance in management of the animals; Dr. P. Pinzani of the Department of Clinical Physiopathology, University of Florence, for the use of the laser microdissector; and Mary Forrest for revision of the English.

1
Caderni G, Femia AP, Giannini A, et al. Identification of mucin-depleted foci in the unsectioned colon of azoxymethane-treated rats: correlation with carcinogenesis.
Cancer Res
2003
;
63
:
2388
–92.
2
Femia AP, Bendinelli B, Giannini A, et al. Mucin-depleted foci have β-catenin gene mutations, altered expression of its protein, and are dose- and time-dependent in the colon of 1,2-dimethylhydrazine-treated rats.
Int J Cancer
2005
;
116
:
9
–15.
3
Femia AP, Caderni G, Bottini C, Salvadori M, Dolara P, Tessitore L. Mucin-depleted foci are modulated by dietary treatments and show deregulation of proliferative activity in carcinogen-treated rodents. Int J Cancer. In press.
4
Femia AP, Dolara P, Caderni G. Mucin-depleted foci (MDF) in the colon of rats treated with azoxymethane (AOM) are useful biomarkers for colon carcinogenesis.
Carcinogenesis
2004
;
25
:
277
–81.
5
Pierre F, Freeman A, Taché S, Van der Meer R, Corpet DE. Beef meat and blood sausage promote azoxymethane-induced mucin-depleted foci and aberrant crypt foci in rat colons.
J Nutr
2004
;
134
:
2711
–6.
6
Yoshimi N, Morioka T, Kinjo T, et al. Histological and immunohistochemical observations of mucin-depleted foci (MDF) stained with Alcian blue, in rat colon carcinogenesis induced with 1,2-dimethylhydrazine dihydrochloride.
Cancer Sci
2004
;
95
:
792
–7.
7
Giles RH, Van Es JH, Clevers H. Caught up in a Wnt storm: Wnt signaling in cancer.
Biochim Biophys Acta
2003
;
1653
:
1
–24.
8
Fearnhead NS, Britton MP, Bodmer WF. The ABC of Apc.
Hum Mol Genet
2001
;
10
:
721
–33.
9
Moser AR, Pitot HC, Dove WF. A dominant mutation that predisposes to multiple intestinal neoplasia in the mouse.
Science
1990
;
247
:
322
–4.
10
Sohn KJ, Puchyr M, Salomon RN, et al. The effect of dietary folate on Apc and p53 mutations in the dimethylhydrazine rat model of colorectal cancer.
Carcinogenesis
1999
;
20
:
2345
–50.
11
De Filippo C, Caderni G, Bazzicalupo M, et al. Mutations of the Apc gene in experimental colorectal carcinogenesis induced by azoxymethane in F344 rats.
Br J Cancer
1998
;
77
:
2148
–51.
12
Bird RP. Observation and quantification of aberrant crypts in the murine colon treated with a colon carcinogen: preliminary findings.
Cancer Lett
1987
;
37
:
147
–51.
13
Pretlow TP, Pretlow TG. Mutant KRAS in aberrant crypt foci (ACF):initiation of colorectal cancer?
Biochim Biophys Acta
2005
;
1756
:
83
–96.
14
European Community. European Community Regulations on the Care and Use of Laboratory Animals, Law 86/609/EC; 1986.
15
Fodde R, Smits R, Clevers H. Apc, signal transduction and genetic instability in colorectal cancer.
Nat Rev Cancer
2001
;
1
:
55
–67.
16
Luchtenborg M, Weijenberg MP, Roemen GM, et al. Apc mutations in sporadic colorectal carcinomas from the Netherlands Cohort Study.
Carcinogenesis
2004
;
25
:
1219
–26.
17
Otori K, Konishi M, Sugiyama K, et al. Infrequent somatic mutation of the adenomatous polyposis coli gene in aberrant crypt foci of human colon tissue.
Cancer
1998
;
83
:
896
–900.
18
Jen J, Powell SM, Papadopoulos N, et al. Molecular determinants of dysplasia in colorectal lesions.
Cancer Res
1994
;
54
:
5523
–6.
19
Takayama T, Ohi M, Hayashi T, et al. Analysis of K-ras, Apc, and β-catenin in aberrant crypt foci in sporadic adenoma, cancer, and familial adenomatous polyposis.
Gastroenterology
2001
;
121
:
599
–611.
20
Van de Wetering M, Sancho E, Verweij C, et al. The β-catenin/TCF-4 complex imposes a crypt progenitor phenotype on colorectal cancer cells.
Cell
2002
;
111
:
241
–50.