Skip to main content
NCBI home page
International Journal of Experimental Pathology logoLink to International Journal of Experimental Pathology
. 2010 Aug;91(4):314–323. doi: 10.1111/j.1365-2613.2010.00704.x

The effect of desalivation on the malignant transformation of the tongue epithelium and associated stromal myofibroblasts in a rat 4-nitroquinoline 1-oxide-induced carcinogenesis model

Marilena Vered 1, Osnat Grinstein-Koren 1, Shoshana Reiter 1, Irit Allon 1, Dan Dayan 1
PMCID: PMC2962890  PMID: 20353426

Abstract

The aim of our study was to analyse desalivated rat tongue epithelium for histopathological changes, proliferating cell nuclear antigen (PCNA), and epithelium-associated stromal myofibroblasts [SMF; α-smooth muscle actin (αSMA)] following 0.001% 4-nitroquinoline 1-oxide (4NQO) administration in drinking water. Results were compared with those of identically treated but salivated specimens. 4NQO was administered for 7, 14, 22 and 28 weeks. Tongue length was divided into anterior, middle and posterior ‘thirds’. The histopathological changes per ‘third’ were scored as normal epithelium, hyperplasia, dysplasia, carcinoma-in-situ, and superficial and invasive carcinoma. The PCNA and αSMA stains were assessed by a point-counting method. At all time points, the histopathological changes in the anterior and middle thirds were higher in the desalivated than in the salivated group (P<0.05) but almost identical in the posterior third (P>0.05). PCNA scores were significantly lower in the desalivated vs. the salivated group at almost all time points and tongue thirds (P < 0.05). SMF were usually scarce in both groups, but there was a significant surge in the posterior third at 28 weeks: the score in the desalivated group was only about one-half that of the salivated group (P<0.05). The absence of saliva seems to promote malignant transformation of the tongue epithelium in the early stages. PCNA cannot be regarded as a marker of proliferation and probably contributes to this process by other mechanisms. Emergence of SMF seems to be highly dependent on growth factors from saliva in addition to factors from cancerous cells.

Keywords: 4NQO tongue cancer, desalivation, PCNA, rats, stromal myofibroblasts

The systemic 4-nitroquinoline 1-oxide (4NQO)-induced tongue carcinogenesis model established in rats has been a source of important information about the evolution of carcinomatous changes that are considered to occur in the same way as they do in humans (Vered et al. 2005). In general, the tongue epithelium underwent clinical changes of increasing severity, ranging from white patches (leukoplakia) to ulcerated, exophytic and papillary tumours (Dayan et al. 1997). The clinical changes lagged behind the nuclear and molecular changes, which could be identified even in ‘normal-looking’ epithelium (Dayan et al. 1997; Ribeiro et al. 2004, 2005, 2007). To assess the role of saliva, the medium by which the carcinogen is distributed to the oral mucosae, and which is regarded as a potent antioxidant milieu (Nagler & Dayan 2006), several studies (Dayan et al. 1997; Kaplan et al. 2001) focused on comparing 4NQO-treated desalivated and salivated rats. The main conclusions were that the frequency of oral lesions increased with time of exposure in all animals, but could be identified substantially earlier, were more frequent and were of a more severe degree in the desalivated rats compared with the salivated group (Dayan et al. 1997). Although changes in the proliferation activity of the tongue epithelium, as assessed by AgNOR staining, were readily observed in all rats even before macroscopic or microscopic changes were detected, they were significantly higher in number and could be identified remarkably earlier among the desalivated rats (Kaplan et al. 2001). All these dissimilarities diminished with time, however, with almost no differences being apparent at the end of the experiment, implying that saliva could have a temporary anti-carcinogenic protective effect, which could both delay and decrease the level of proliferation induced by the carcinogen.

In a more recent study, we also focused on stromal myofibroblasts (SMF) in rats with normally functioning salivary glands and found that an increased density of SMF was distinctively associated with the development of carcinoma but not with the premalignant changes that were observed within the epithelium (Vered et al. 2007). Ultrastructurally, these carcinoma cells revealed unique changes in the organization of the basal lamina and the presence of cytoplasmic microfilaments compatible with contractile fibres (Vered et al. 2008). We assumed that this could reflect essential changes in the phenotype of the carcinoma cells towards a mesenchymal differentiation.

This study aimed to assess changes in histomorphology and proliferation activity (proliferating cell nuclear antigen, PCNA) of the tongue epithelium, as well as in the density of the epithelium-associated SMF (α-smooth muscle actin; αSMA) in desalivated rats and to compare those results to the salivated and control (normal) groups. We hoped to clarify further the role of saliva in the process of development of tongue carcinoma in the rat model, which has possible implications for human patients.

Methods

The ‘desalivated’ group comprised 39 rats described in detail in the study of Dayan et al. (1997) in which they were designated as the ‘experimental’ group. The rats in this group underwent desalivation procedures by ligation of the parotid salivary glands and removal of the submandibular and sublingual glands, and were administered 4NQO dissolved in tap water to a final concentration of 0.001%. Carcinogen administration lasted for 7 (n=4), 8 (n=7), 14 (n=7), 22 (n=8) and 28 weeks (n=13), at the end of which rats were killed by CO2. As a result of the relatively small number of animals in the first time point and the general trend of a lack of any significant differences between the animals in the first two time points – unlike the findings in the original study – the rats in these two groups were combined in the current study (T1). The remaining time points were therefore marked as T2, T3 and T4. The ‘salivated’ group of rats in this study was the ‘control group’ in our earlier study (Dayan et al. 1997): they underwent a sham operation that left their salivary glands fully functional. To perform a more comprehensive comparison between the results of the desalivated and salivated groups, we also combined the first and second time points (i.e. 7 and 8 weeks of 4NQO administration) in the salivated group and formed a new group, T1C (n=12), where the suffix ‘C’ represents the salivated (control) group. Accordingly, the remaining time points of the salivated group were designated as T2C (n=6), T3C (n=9) and T4C (n=9) and conformed to the time points of the desalivated group. We used yet another group of rats, ‘normal’ group (T0, n=7, no operation, no carcinogen administration) that had been part of our previous study (Vered et al. 2007) for purposes of further comparison with the desalivated rats.

The tongues were dissected, fixed in formalin and embedded in paraffin. One haematoxylin and eosin slide and two slides for immunohistochemical stains were prepared from each block. The haematoxylin and eosin-stained slide served to assess the severity of the cytological and morphological changes within the tongue epithelium. The histopathological assessment of the degrees of the dysplastic changes for all the experimental groups were performed separately for each ‘third’ of the epithelial lining, unlike the initial work on the same series of animals in which the entire dorsal surface epithelium was assessed for the worst changes (Dayan et al. 1997). In this study, similar to the procedure we followed in a previous study (Vered et al. 2007), the distance between the tip of the tongue to its most posterior aspect was marked on a glass slide and divided into three equal parts, designating the ‘anterior’, ‘middle’ and ‘posterior’ thirds. The rational for this partition lies in the fact that it can show the entire spectrum of changes in an increasing order of severity not only as a factor of time of exposure to the carcinogen, but also as a factor of location along the tongue. In this way, histopathological features of even ‘normal-looking’ epithelium can be examined vis-à-vis the corresponding molecular changes. In addition, this supplies a consistent and repeatable methodology, which should enable a comprehensive comparison among similar studies.

The histopathological changes were classified for each third on an 8-point scale for scoring changes ranging from normal-looking epithelium (a score of 0), hyperplasia and/or hyperkeratosis (a score of 1), increasing severity of dysplasia (scores of 2, 3 and 4 for mild, moderate and severe dysplasia respectively), carcinoma in situ (a score of 5), superficial carcinoma (a score of 6) and invasive carcinoma (a score of 7). Results were expressed as the mean score of the histopathological changes for each ‘third’ and for the time point for the desalivated, salivated and normal groups.

The other slides were used to quantify the proliferative potential of the epithelial cells (PCNA, clone PC-10, 1:800, Zymed, San Francisco, CA, USA) and the frequency of the SMF (αSMA, clone 1A4, 1:100, Dako A/S, Denmark). The immunomorphometric assessment of the PCNA-stained epithelial cells and αSMA-stained SMF in this study groups followed the guidelines previously established for the salivated group (Vered et al. 2007). The results were presented as the mean per cent of the stained cells per intersection per each ‘third’ and the time point for the desalivated, salivated and normal groups.

This study was approved by the Animal Committee of the Sackler Faculty of Medicine, Tel Aviv University and conformed to procedures described in the Guiding Principles for the use of Laboratory Animals.

Statistical analysis

The pattern of changes in each parameter among the tongue ‘thirds’ in the study groups was analysed using anova with repeated measures. Differences in the mean scores for each parameter between a given study group and the salivated group per ‘third’ at each time point were analysed using the t-test. The impact of the presence of saliva and length of exposure to the carcinogen on each parameter was assessed using two-way anova. Differences were considered significant at P<0.05. The statistical calculations were performed using the spss software, version 16 (SPSS Inc., Chicago, IL, USA).

Results

Histopathological diagnoses

A similar trend of an increase in the mean scores from the anterior to the middle and then to the posterior third was observed at all time points in the desalivated group. The values increased as a factor of the length of exposure to the carcinogen. The highest scores were consistently found in the T4 group in all ‘thirds’ (P<0.001) (Figure 1). Even the anterior third had marked changes corresponding to the range of severe dysplasia to carcinoma in situ. The mean scores in the normal group were negligible and did not change among tongue ‘thirds’: they were significantly lower than those of the other ‘thirds’ (0 in the anterior third, 0.5 ± 1 in the middle and 0 ± 1 in the posterior third) (P<0.001).

Figure 1.

Figure 1

Open in a new tab

Histopathological scores in the desalivated and salivated groups according to tongue thirds and time points. Ant, anterior third; Mid, middle third; Post, posterior third; HP + HK, hyperplasia and hyperkeratosis; Mod dysplasia, moderate dysplasia; Sev dysplasia, severe dysplasia; CIS, carcinoma in situ; Sup SCC, superficial squamous cell carcinoma; Inv SCC, invasive squamous cell carcinoma; NS, no significance.

A comparison of the results between the desalivated and salivated groups (Figure 1) showed that the mean scores for the anterior ‘third’ were very similar (0.9 ± 0.7 and 0.4 ± 0.5 respectively) and corresponded to hyperplasia and hyperkeratosis in both groups at T1 (P>0.05). At T2, the desalivated group had a mean score approaching moderate dysplasia (2.8 ± 0.4), whereas the salivated group had a mean score indicative of less than mild dysplasia (1.5 ± 0.8). The discrepancy between the groups continued to increase in parallel to the continuing administration of the carcinogen, so that the mean score of the desalivated group at T4 approximated carcinoma in situ (4.6 ± 1.6) while that of the salivated group was less than moderate dysplasia (2.6 ± 0.6) (P-values in Figure 1). The same trend of changes in the histopathological diagnoses and differences between the desalivated and salivated groups were observed in the middle ‘third’. In particular, the mean score in the desalivated group at T4 was almost superficial carcinoma (5.8 ± 1.3) while that in the salivated group was almost moderate dysplasia (3.1 ± 1.1) (P-values in Figure 1). The posterior ‘third’ exhibited a completely different picture compared to both the anterior and middle ‘thirds’ in that both the desalivated and the salivated groups had very similar or almost identical mean values at all time points. Specifically, both groups had mean scores corresponding to a range of hyperplasia-to-mild dysplasia at T1 (1.4 ± 1.0 for the salivated and 2.1 ± 1.8 for the desalivated group respectively), moderate-to-severe dysplasia at T2 (3.2 ± 0.7 for the salivated and 4 ± 1 for the desalivated group respectively), superficial carcinoma at T3 (5.7 ± 1.1 and 5.9 ± 0.8 for the salivated and desalivated groups respectively) and invasive carcinoma at T4 (6.9 ± 0.3 and 7 ± 0 for the salivated and desalivated groups respectively) (P>0.05).

PCNA mean scores

The pattern of changes that was observed at all time points in the desalivated group was similar for the anterior to the middle ‘thirds’ and showed a mild increase in the mean PCNA scores (Figure 2). The mean scores of the middle to the posterior ‘thirds’ continued to increase at T1, T3 and T4 but decreased at T2 (P=0.038). The mean PCNA scores for all ‘thirds’ in the normal group were significantly higher than the scores at all other thirds (23.2 ± 8.5 in the anterior third, 21 ± 4.3 in the middle and 24.3 ± 4.1 in the posterior third) (P<0.001).

Figure 2.

Figure 2

Open in a new tab

Proliferating cell nuclear antigen scores in the desalivated and salivated groups according to tongue thirds and time points. Ant, anterior third; Mid, middle third; Post, posterior third; HP, hyperplasia; Mild dys, mild dysplasia; Mod dys, moderate dysplasia; Sev dys, severe dysplasia; CIS, carcinoma in situ; Sup SCC, superficial squamous cell carcinoma; Inv SCC, invasive squamous cell carcinoma; NS, no significance.

The mean PCNA scores for the desalivated and salivated groups were compared in parallel with the histopathological diagnoses corresponding to the mean scores obtained at each time point in each group (Figure 2). There was a general trend wherein the mean PCNA scores in the desalivated group were significantly lower than those in the salivated group at all ‘thirds’ and at all time points, with the exception of T2-T2C in the anterior (10.9 ± 3.7 and 15.4 ± 5.5 respectively) and middle (13.4 ± 3.3 and 18 ± 5.4 respectively) ‘thirds’; however, the histopathological diagnoses were always of a more severe degree in the desalivated than in the salivated group. The most notable difference in the mean PCNA scores was found in the posterior ‘third’ at T4-T4C (15.8 ± 3.9 and 46.6 ± 15 respectively), although both groups were assessed histopathologically as being invasive squamous cell carcinoma (SCC) (P-values in Figure 2). Figure 3 illustrates the pattern of PCNA-stained tumours in rats representing the desalivated and salivated groups.

Figure 3.

Figure 3

Open in a new tab

Proliferating cell nuclear antigen (PCNA) in carcinomas from desalivated (a) and salivated (b) rats. The PCNA-positive nuclei are usually arranged in one layer at the periphery of the tumours or even absent (arrow) in the former, while in the latter the positive nuclei are commonly present throughout the tumour (ABC method, ×100 original magnification).

αSMA mean scores

The mean scores in the anterior ‘third’ at all time points were low and did not exceed one stained cell per intersection in the desalivated group (Figure 4). In the middle ‘third’, the mean scores at T1, T2 and T3 were within the same range as that of the anterior ‘third’, while there was a slight increase towards a mean score of 1.5 stained cells per intersection at T4. The mean αSMA scores in the posterior ‘third’ remained essentially unchanged at T1 and T2 (about a mean score of 1). There was an increase in the mean score to slightly above two stained cells per intersection in T3, while the most remarkable change in the posterior ‘third’ was found at T4, where the mean score showed more than a two-fold increase compared with the middle ‘third’ (P=0.008). The mean scores of the αSMA-stained cells in the normal group (T0) was low in all tongue ‘thirds’ (0.9 ± 0.5 in the anterior, 1 ± 1 in the middle and 0.8 ± 0.8 in the posterior third), and were only significantly lower than the desalivated group at T4 in the posterior ‘third’ (P<0.006).

Figure 4.

Figure 4

Open in a new tab

α-Smooth muscle actin scores in the desalivated and salivated groups according to tongue thirds and time points; NS, no significance. Ant, anterior third; Mid, middle third; Post, posterior third.

The patterns observed in the anterior and middle ‘thirds’ were similar for the desalivated and salivated groups (Figure 4). In general, the mean values were low and the mean scores in the desalivated group were significantly lower than those in the salivated group at T1 and T2 (P-values in Figure 4), but they were similar at T3 and T4 (P>0.05). In the posterior ‘third’, the mean scores remained low and there were no significant differences between the desalivated and salivated groups at T1-T1C to T3-T3C. At T4-T4C, however, there was a remarkable surge in the mean scores in both groups, although the increase in the desalivated group (3.9 ± 3.3) was only about one-half of that in the salivated group (8.5 ± 4) (P=0.011). Figure 5 illustrates the pattern of αSMA-stained tumours in rats representing the desalivated and salivated groups.

Figure 5.

Figure 5

Open in a new tab

α-Smooth muscle actin (αSMA) in carcinomas from desalivated (a) and salivated (b) rats. In the former the αSMA-positive stromal fibroblasts occur in small numbers, while in the latter they encircle the tumour islands abundantly (ABC method, ×200 original magnification).

Impact of the presence of saliva and the length of time of exposure to the carcinogen on the examined parameters

The different patterns of changes in the histopathological diagnoses observed with time of exposure to the carcinogen in the anterior and middle ‘thirds’ were associated with the status of the saliva (P=0.034 and P=0.014 respectively). For PCNA, the changes in the mean scores as a function of time of exposure were associated with the status of the saliva in the middle and posterior ‘thirds’ (P=0.024 and P=0.013 respectively). For αSMA, the changes in the mean scores at the different time points were associated with the status of the saliva in the posterior ‘third’ (P=0.001).

Discussion

This study assessed the pattern of 4NQO-induced changes in the epithelial histomorphology, PCNA signal and SMF frequency in the tongues of desalivated rats according to tongue ‘thirds’ and to time points of carcinogen administration. We found that the general pattern of the histomorphological changes was that of increasing severity in the epithelial pathology, most markedly in the posterior ‘third’, as the animals were exposed to a carcinogen for longer periods of time. The occurrence of PCNA-positive cells showed a slight tendency to increase in the posterior location, with only very slight differences in terms of length of carcinogen administration. The SMF pattern was notable for consistently low values at the different time points and tongue locations, with the exception of the posterior ‘third’, among the animals that were given a carcinogen for the longest period of time (28 weeks, T4): there was a surge in the number of SMF at T4. The present group of desalivated rats had been previously assessed for histopathological changes (Dayan et al. 1997). In that study, an analysis was performed on the entire length of the dorsal tongue epithelium, and a general trend of progressive epithelial pathologies was observed as a factor of the period of time of carcinogen administration. In this study, which employed a methodology of assessment per tongue ‘thirds’, we were able to show that the evolution of the severity of the histopathological changes depended on both the length of time of exposure and the location along the tongue. Although only one case of invasive SCC developed in the posterior ‘third’ in the desalivated group at T1 [also identified clinically (Dayan et al. 1997)], both the anterior and middle ‘thirds’ exhibited histopathological changes with obvious malignant potential, emphasizing the principle of field cancerization (Slaughter et al. 1953).

A comparison of the histopathological changes between the salivated and desalivated groups revealed that both groups had a similar degree of changes in the anterior and middle ‘thirds’ at T1. With a longer duration of carcinogen administration, however, there was a significantly increase in divergence of the mean scores: the desalivated group always had the higher scores. By contrast, irrespective of the length of carcinogen administration, the mean scores of the histopathological changes were almost identical in the posterior ‘third’ in both the desalivated and salivated groups. This observation provides more specific evidence for our earlier, more generalized conclusion that there were almost no differences between the desalivated and salivated groups at the end of the experiment (Dayan et al. 1997). The water soluble 4NQO (Ohne et al. 1985) that we used is expected to be diluted by the saliva, thus accounting for its lower dysplastic effect in the salivated group, especially in the anterior and middle ‘thirds’. On the other hand, carcinogenic agents, such as those originating in cigarette smoke/tobacco, are assumed not only to neutralize the antioxidant potential of the saliva but also to bring about the generation of deleterious oxidative DNA damage from redox-active metallic compounds (i.e. iron and copper) secreted from the major salivary glands (Nagler & Dayan 2006). The facts that rats drink in such a way that the ingested liquid concentrates in the posterior region of the tongue and that saliva plays a role in the pathogenesis of SCC in the presence of a carcinogenic agent (Nagler & Dayan 2006) can serve to explain the high histopathological scores in the posterior tongue in the salivated group. Although the rats in the desalivated group might have consumed less of the carcinogen-containing water as a result of their deteriorating physical condition (Dayan et al. 1997), the carcinogen is more concentrated and therefore induces more severe lesions along the entire dorsal epithelium and, in particular, on the posterior ‘third’.

Our current PCNA results yielded consistently significant lower values in the desalivated group compared with the salivated group. In the latter group, the increase in the PCNA scores as a factor of length of exposure to the carcinogen was in accordance with a similar study, in which it was nicely shown that PCNA was closely associated with a high cellular proliferation activity during the neoplastic transformation (Silva et al. 2007). The rationale for detecting a purportedly lower signal of proliferation in the desalivated group apparently lies in the observation that the lack of saliva deprives the oral epithelial milieu of important potential growth factors, i.e. epidermal growth factor (EGF) and nerve growth factor (NGF) that are produced in and secreted by the salivary glands (Mathison et al. 2004). Considering that the histopathological scores in the desalivated group were always higher, thus implying higher proliferation activity, these findings seem to be contradictory. One feasible explanation can be provided by the accumulating evidence that PCNA should not be regarded as a mere marker of proliferation, but rather as a factor that stands at the very heart of many essential cellular processes, such as DNA replication, repair of DNA damage, chromatin structure maintenance, chromosome segregation and cell-cycle progression (Garg & Burgers 2005; Brown et al. 2009; Stoimenov & Helleday 2009). If DNA contains damaged foci when the replication begins, the replication fork may encounter that damage, stall and eventually collapse. Stalled replication forks are the signal for activation of special pathways involved in the avoidance of DNA damage, which are believed to be controlled through interactions with PCNA; however, the key step seems to be a post-translational modification of the PCNA molecule. The current knowledge suggests that at the stalled replication forks, PCNA becomes mono-ubiquitylated on an evolutionarily conserved position (a lysine residue at position 164). This is a signal for recruitment of a special polymerase, which is able to continue DNA replication even on a damaged template. The antibody used for the immunohistochemical detection of PCNA in this study was the frequently used PC10, whose epitope is in the vicinity of the ubiquination region (Roos et al. 1993). In light of this, it would be reasonable to assume that ubiquination of the PCNA molecule could give rise to a spatial interference of the interaction between the antibody and its molecular epitope, hence the lower PCNA signal received from the specimens of the desalivated rats. Once the DNA replication is renewed after putting the damage-avoidance ‘plan’ into operation, these cells begin to accumulate a growing number of genetic aberrations, which ultimately lead to a frequent and more conspicuous occurrence of dysplastic changes and full-blown carcinomatous transformation. As such, this assumption is in accordance with the histomorphological results of the desalivated group. In the salivated group, where (i) the epithelial environment contained EGF and NGF and (ii) the carcinogen was more diluted, it can be suggested that the resulting conditions promoted PCNA activity either towards replication or DNA repair, which could have readily been identified by the immunohistochemical staining, as reflected by the high PCNA scores in this group.

SMF have been recognized in recent years as being one of the main tumour-promoting partners of carcinomas (Eiden 2007). Their presence in the close vicinity of the malignant cells is induced by factors produced by them, such as platelet-derived growth factor and transforming growth factor-β (TGF-β) (De Wever & Mareel 2003). In our previous study on the salivated group of the current study, the stroma adjacent to the epithelium, which underwent gradual dysplastic changes, was almost devoid of SMF up to the moment it was fully transformed, at which point abundant SMF emerged (Vered et al. 2007). The present changes in SMF in the desalivated group followed the pattern described in the salivated group on a qualitative but not a quantitative level: the increase in the SMF that accompanied the carcinomas in the former was only about one-half of that in the latter. Several recent studies have demonstrated the existence of interactions between EGF and TGF-β in the promotion of differentiation of SMF. He and Bazan (2008) demonstrated synergism between EGF and TGF-β that increase the differentiation and migration of myofibroblasts in the cornea. Under culture conditions, TGF-β induced 12% of the cells to differentiate into myofibroblasts, while EGF plus TGF-β stimulated 90% of cells to differentiate into myofibroblasts. Narine et al. (2006) investigated other growth factors in combination with TGF-β to stimulate the proliferation of suitably differentiated dermal mesenchymal cells (in vitro) and to enhance their invasion into aortic valve matrices. They showed that the combination of TGF- β and basic fibroblast growth factor (bFGF) and EGF upregulated αSMA expression in these cells, but that the combination of TGF-β and EGF was a significantly stronger inducer of the myofibroblast phenotype (αSMA expression) than combined TGF-β and bFGF. As both EGF and TGF-β are produced within the salivary glands and are present in the saliva (Nakabayashi et al. 2001; Rezaie et al. 2006; Yousefzadeh et al. 2006), they were expected to exert their impact on the formation of the SMF in the salivated group of this study. This is in addition to the effect of the carcinoma-derived TGF-β on the emergence of SMF. In the absence of saliva and its growth factors in the desalivated group, the SMF are probably only the net product of the carcinomatous component. This could be a feasible explanation of the present finding that SMF were less abundant in the desalivated group compared with the salivated one. Wound healing of the oral mucosa and alveolar bone in another animal model of desalivated rats was shown to occur at a slower rate and to involve significantly fewer myofibroblasts than the salivated group (Bodner et al. 1992; Dayan et al. 1992), further supporting the notion that saliva-derived EGF and TGF-β have an important role in the formation of SMF.

In summary, this study analysed the 4NQO-associated pattern of changes in a number of parameters in both the epithelial and stromal components in desalivated and salivated rats. The results highlighted the existence of some complex interactions between the carcinogen, the saliva and its growth factors, the epithelial response and the stromal reaction which had not been investigated thus far. The salivated group of this rat model was assumed to simulate the process of oral carcinogenesis in humans, while the findings in the desalivated group could be regarded as a simulator of the postirradiation end-product of salivary gland dysfunction and, therefore, serve as the starting point of future studies. Although the rat model is considered to differ in some aspects from what actually happens to human patients in postradiation conditions (Grundmann et al. 2009), it is still a valuable tool in terms of understanding the principles of certain biological mechanisms with potential for clinical application.

Acknowledgments

The study was supported by the Alpha Omega Fund, School of Dental Medicine, Tel Aviv University. The authors would like to thank Ms Hana Vered for technical assistance and Ms Esther Eshkol for editorial assistance.

References

  1. Bodner L, Dayan D, Oberman M, Hirshberg A, Tal H. Healing of experimental wounds in sialectomized rat. J. Clin. Periodontol. 1992;19:345–347. doi: 10.1111/j.1600-051x.1992.tb00657.x. [DOI] [PubMed] [Google Scholar]
  2. Brown S, Niimi A, Lehmann A. Ubiquitination and deubiquitination of PCNA in response to stalling of the replication fork. Cell Cycle. 2009;8:689–692. doi: 10.4161/cc.8.5.7707. [DOI] [PubMed] [Google Scholar]
  3. Dayan D, Bodner L, Horowitz I. Effect of salivary gland hypofunction of the healing of extraction wounds: a histomorphometric study in rats. J. Oral Maxillofac. Surg. 1992;50:354–358. doi: 10.1016/0278-2391(92)90397-i. [DOI] [PubMed] [Google Scholar]
  4. Dayan D, Hirshberg A, Kaplan I, Rotem N, Bodner L. Experimental tongue cancer in desalivated rats. Oral Oncol. 1997;33:105–109. doi: 10.1016/s0964-1955(96)00048-6. [DOI] [PubMed] [Google Scholar]
  5. De Wever O, Mareel M. Role of tissue stroma in cancer cell invasion. J. Pathol. 2003;200:429–447. doi: 10.1002/path.1398. [DOI] [PubMed] [Google Scholar]
  6. Eiden B. The Myofibroblast. A Study of Normal, Reactive and Neoplastic Tissues with Emphasis on Ultrastructure. Siena: Nuova Immagine Editrice Press; 2007. [PubMed] [Google Scholar]
  7. Garg P, Burgers PM. Ubiquitinated proliferating cell nuclear antigen activates translesion DNA polymerases and REV1. PNAS. 2005;102:18361–18366. doi: 10.1073/pnas.0505949102. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Grundmann O, Mitchell GC, Limesand KH. Sensitivity of salivary glands to radiation: from animal models to therapies. J. Dent. Res. 2009;88:894–903. doi: 10.1177/0022034509343143. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. He J, Bazan HE. Epidermal growth factor synergism with TGF-beta1 via PI-3 kinase activity in corneal keratocyte differentiation. Invest. Ophthalmol. Vis. Sci. 2008;49:2936–2945. doi: 10.1167/iovs.07-0900. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Kaplan I, Engelberg J, Dayan D. The effect of desalivation on 4-nitroquinoline-1-oxide-induced carcinogenesis: a morphometric study of nucleolar organizer regions. J. Oral Pathol. Med. 2001;30:48–52. doi: 10.1034/j.1600-0714.2001.300108.x. [DOI] [PubMed] [Google Scholar]
  11. Mathison R, Davidson JS, Befus AD. Neuroendocrine regulation of inflammation and tissue repair by submandibular gland factors. Immunol. Today. 2004;15:527–532. doi: 10.1016/0167-5699(94)90209-7. [DOI] [PubMed] [Google Scholar]
  12. Nagler R, Dayan D. The dual role of saliva in oral carcinogenesis. Oncology. 2006;71:10–17. doi: 10.1159/000100445. [DOI] [PubMed] [Google Scholar]
  13. Nakabayashi T, Sakata KM, Sakata A, et al. TGF-beta1 null mutation leads to CD154 upregulated expression in affected tissues. Inflammation. 2001;25:69–73. doi: 10.1023/a:1007158219974. [DOI] [PubMed] [Google Scholar]
  14. Narine K, De Wever O, Van Valckenborgh D, et al. Growth factor modulation of fibroblast proliferation, differentiation, and invasion: implications for tissue valve engineering. Tissue Eng. 2006;12:2707–2716. doi: 10.1089/ten.2006.12.2707. [DOI] [PubMed] [Google Scholar]
  15. Ohne M, Satoh T, Yamada S, Takai H. Experimental tongue carcinoma of rats induced by oral administration of 4-nitroquinoline-1-oxide (4NQO) in drinking water. Oral Surg. Oral Med. Oral Pathol. 1985;59:600–607. doi: 10.1016/0030-4220(85)90189-6. [DOI] [PubMed] [Google Scholar]
  16. Rezaie A, Ghorbani F, Eshghtork A, et al. Alterations in salivary antioxidants, nitric oxide, and transforming growth factor-beta 1 in relation to disease activity in Crohn’s disease patients. Ann. N. Y. Acad. Sci. 2006;1091:110–122. doi: 10.1196/annals.1378.060. [DOI] [PubMed] [Google Scholar]
  17. Ribeiro DA, Fávero Salvadori DM, da Silva RN, Ribeiro Darros B, Alencar Marques ME. Genomic instability in non-neoplastic oral mucosa cells can predict risk during 4-nitroquinoline 1-oxide-induced rat tongue carcinogenesis. Oral Oncol. 2004;40:910–915. doi: 10.1016/j.oraloncology.2004.04.010. [DOI] [PubMed] [Google Scholar]
  18. Ribeiro DA, Salvadori DM, Marques ME. Abnormal expression of bcl-2 and bax in rat tongue mucosa during the development of squamous cell carcinoma induced by 4-nitroquinoline 1-oxide. Int. J. Exp. Pathol. 2005;86:375–381. doi: 10.1111/j.0959-9673.2005.00444.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Ribeiro DA, Kitakawa D, Domingues MA, Cabral LA, Marques ME, Salvadori DM. Survivin and inducible nitric oxide synthase production during 4NQO-induced rat tongue carcinogenesis: a possible relationship. Exp. Mol. Pathol. 2007;83:131–137. doi: 10.1016/j.yexmp.2007.02.006. [DOI] [PubMed] [Google Scholar]
  20. Roos G, Landberg G, Huff JP, Houghten R, Takasaki Y, Tan EM. Analysis of epitopes of proliferating cell nuclear antigen recognized by monoclonal antibodies. Lab. Invest. 1993;68:204–210. [PubMed] [Google Scholar]
  21. Silva RN, Ribeiro DA, Salvadori DM, Marques ME. Placental glutathione S-transferase correlates with cellular proliferation during rat tongue carcinogenesis induced by 4-nitroquinoline 1-oxide. Exp. Toxicol. Pathol. 2007;59:61–68. doi: 10.1016/j.etp.2007.02.010. [DOI] [PubMed] [Google Scholar]
  22. Slaughter DP, Southwick HW, Smejkal W. Field cancerization in oral stratified squamous epithelium; clinical implications of multicentric origin. Cancer. 1953;6:963–968. doi: 10.1002/1097-0142(195309)6:5<963::aid-cncr2820060515>3.0.co;2-q. [DOI] [PubMed] [Google Scholar]
  23. Stoimenov I, Helleday T. PCNA on the crossroad of cancer. Biochem. Soc. Trans. 2009;37:605–613. doi: 10.1042/BST0370605. [DOI] [PubMed] [Google Scholar]
  24. Vered M, Yarom N, Dayan D. 4NQO oral carcinogenesis: animal models, molecular markers and future expectations. Oral Oncol. 2005;41:337–339. doi: 10.1016/j.oraloncology.2004.07.005. [DOI] [PubMed] [Google Scholar]
  25. Vered M, Allon I, Buchner A, Dayan D. Stromal myofibroblasts and malignant transformation in a 4NQO rat tongue carcinogenesis model. Oral Oncol. 2007;43:999–1006. doi: 10.1016/j.oraloncology.2006.11.007. [DOI] [PubMed] [Google Scholar]
  26. Vered M, Polak-Charcon S, Babushkin T, Dayan D. 4NQO-induced rat tongue carcinoma: an ultrastructural study. Ultrastruct. Pathol. 2008;32:199–205. doi: 10.1080/01913120802034645. [DOI] [PubMed] [Google Scholar]
  27. Yousefzadeh G, Larijani B, Mohammadirad A, et al. Determination of oxidative stress status and concentration of TGF-beta 1 in the blood and saliva of osteoporotic subjects. Ann. N. Y. Acad. Sci. 2006;1091:142–150. doi: 10.1196/annals.1378.062. [DOI] [PubMed] [Google Scholar]

Articles from International Journal of Experimental Pathology are provided here courtesy of Wiley

RESOURCES