The effects of short-chain fatty acids on colon epithelial proliferation and survival depend on the cellular phenotype
- First Online:
- Cite this article as:
- Comalada, M., Bailón, E., de Haro, O. et al. J Cancer Res Clin Oncol (2006) 132: 487. doi:10.1007/s00432-006-0092-x
- 424 Views
Purpose: The short-chain fatty acids (SCFA) are produced via anaerobic bacterial fermentation of dietary fiber within the colonic lumen. Among them, butyrate is thought to protect against colon carcinogenesis. However, few studies analyze the effects of butyrate, and other SCFA, on normal epithelial cells and on epithelial regeneration during disease recovery. Since there are controversial in vitro studies, we have explored the effects of SCFA on different biological processes. Methods: We used both tumoral (HT-29) and normal (FHC) epithelial cells at different phenotypic states. In addition, we analyzed the in vivo activity of soluble dietary fiber and SCFA production in the proliferation rate and regeneration of intestinal epithelial cells. Results: The effect of butyrate on epithelial cells depends on the phenotypic cellular state. Thus, in nondifferentiated, high proliferative adenocarcinoma cells, butyrate significantly inhibited proliferation while increased differentiation and apoptosis, whereas other SCFA studied did not. However, in normal cells or in differentiated cultures as well as in in vivo studies, the normal proliferation and regeneration of damaged epithelium is not affected by butyrate or SCFA exposure. Conclusion: Although butyrate could exert antiproliferative effects in tumor progression, its production is safe and without consequences for the normal epithelium growth.
KeywordsTumor cellsColorectal cancerFiberFermentationColonic damage
Intestinal epithelial cells (IEC) constitute a single layer that is folded to form a number of invaginations, or crypts, which are embedded in the connective tissue. Crypt size and organization are generally uniform within a given region of the gastrointestinal tract (Potten 1995). Intestinal stem cells reside at the origin of the migration, which is found just above the crypt base in the small intestine and at the crypt base in the colon. Hence, the differentiated, functional cells are found mainly on the villi (small intestine) or toward the top of the colonic crypt, the intercrypt table, in the large intestine. During the latest stages of the differentiation process, these mature epithelial cells become senescent and are shed intact into the lumen. Cells shed from the gut must be replaced by a steady supply of cells generated in the low-to-mid-crypt region (Kaur and Potten 1986).
In addition to their classical absorptive and physical barrier roles, IEC are now viewed as immunological sentinels of the gut (Kagnoff and Eckmann 1997). Similarly to normal homeostatic proliferation, crypt regeneration after cytotoxic damage also appears to be initiated in the crypts. In several experimental models of colitis, it has been reported that epithelial cell proliferation is increased in association with the inflammatory process (Ioachim et al. 2004; Weiss et al. 2004). However, it is unclear whether this proliferative activity is due to direct stimulation of cell cycling by cytokines, growth factors, and other mediators or whether it is a response to epithelial cell loss. In consequence, enhanced proliferation may be a compensatory response to replace the lost cells due to necrosis or apoptosis.
Unfortunately, the increase of cell turnover in these stem cells, responsible for tissue homeostasis and regeneration, can enhance the risk of mutations leading to a carcinogenic phenotype (Vogelstein et al. 1988). It is also thought that defective apoptosis may play a role in carcinogenesis since certain genes that are associated with the development of gastrointestinal cancer (p21Waf1, p53, and Bcl-2) are known to regulate apoptosis and proliferation (El-Deiry et al. 1995; Merritt et al. 1995, 1997).
Colorectal cancer is one of the most common solid tumors worldwide with high metastatic potential. Clinical evidences showed that the removal of colorectal adenoma could attenuate 76–90% risk of colorectal cancer, with a yearly relapse rate of about 10–15%. For this reason, the urgent task is to develop new strategies to prevent the disease. Evidences have shown that low fat and high dietary fiber diet could protect against colorectal cancer (Hinnebusch et al. 2002; Zhang et al. 2004).
Dietary fiber fermentation by the colonic bacterial flora produces short-chain fatty acids (SCFA), mainly acetate, propionate, and butyrate, in nearly constant molecular ratio of 1, 0.3, and 0.25, respectively (D’Argenio and Mazzacca 1999). Among them, butyrate is considered as the major energy substrate for colonocytes, and at least in rats, it seems to protect against colonic carcinogenesis (Young et al. 1996), probably related to its ability to induce inhibition of histones phosphorylation and thus impairment of cellular growth (Whitlock et al. 1980). Hence, butyrate (or dietary fiber) seems a good candidate for cancer treatment for its antiproliferative function. However, acetate and propionate are studied in a lower rate and their implications on the physiology of the colonocytes are not well documented.
In addition, it has also been suggested that butyrate plays an important role in mucosal repair in inflammatory conditions (Roediger 1980; Scheppach et al. 1992). Thus, different studies have shown that topical SCFA treatment can effectively be used to treat distal ulcerative colitis, diversion colitis, nonspecific protosigmoiditis in humans (Breuer et al. 1991; Harig et al. 1989; Scheppach et al. 1992; Senagore et al. 1992), and experimental bowel inflammation in rats (Butzner et al. 1996; D’Argenio et al. 1994). In consequence, butyrate (or other SCFA) could actively participate in epithelium regeneration induced after colonic damage.
Considering the above, which function really exerts butyrate and the other SCFA on epithelial homeostasis? Are they involved in the antiproliferative effects of dietary fiber responsible for the protection against colonic cancer? Or, in contrast, are they required during epithelial regeneration during disease recovery? Finally, what is the effect of SCFA when they are produced during the normal proliferation rate of the epithelial cells?
In this study, we have tried to answer some of these questions by analyzing the in vitro effects of SCFA in both cancer and normal intestine epithelial cells. Also, we studied the effects of a soluble dietary fiber, capable to promote colonic SCFA production, in the proliferation rate and regeneration of IEC in an experimental model of rat colitis, in which the regeneration of epithelial cells is a crucial process.
The results obtained showed that the effects of butyrate on cell proliferation rate depend on the cellular differentiation status. All SCFA do not affect the proliferation, viability, and differentiation status in normal epithelial cells (FHC) nor in low-proliferating differentiated adenocarcinoma cells (HT-29), which may represent the normal epithelium barrier. However, butyrate induces differentiation and cell death in active proliferating HT-29. In accordance with these results, SCFA mix (butyrate, propionate, and acetate) naturally produced by fructo-oligosaccharides (FOS) fermentation in rat colonic lumen, significantly affects HT-29 proliferating cells without showing a significant effect in HT-29 differentiated culture, and it does not impair the proliferation process required during in vivo regeneration of a damaged intestinal mucosa. In conclusion, we give some clues to explain the controversial in vitro studies about epithelial cells, since the phenotypic status of these cells exerts different biological functions.
Materials and methods
All reagents and culture media, unless otherwise stated, were purchased from Sigma (Madrid, Spain).
The human colon adenocarninoma cells (HT-29) and fetal human normal colon cells (FHC) were obtained from the Cell Culture Unit of the University of Granada (Granada, Spain), and were grown following ATCC recommendations. They were passaged twice a week after exposure to trypsin/EDTA (0.05/0.02%) with a plating density of 1:3 (5–6×105 cells/ml). For the experiments, cells were seeded in 6- or 24-well plates, depending on the degree of confluence required. For the differentiated culture experiments, confluent HT-29 cells were maintained during 18 days.
Cell proliferation of both cell lines was assessed by Thymidine incorporation assay previously described for other cells (Xaus et al. 2001) with some modifications. 1×105 cells (nonconfluent), 5×105 cells (subconfluent), 15×105 cells (confluent), and differentiated culture were incubated for 24 h in 24-well plates in 1 ml/well of complete medium in the presence or absence of the indicated SCFA concentrations. In the case of FHC cells, the plates were cultivated until subconfluence. Then, 3H-thymidine (1 μCi/ml) was added (ICN Pharmaceuticals Inc., Costa Mesa, CA, USA). After 6 h of incubation at 37°C in the HT-29 culture cells and 24 h in FHC cells, the media were removed and the cells were fixed in ice-cold 70% methanol. After three washes in ice-cold TCA, the cells were solubilized in 1% SDS and 0.3 M NaOH at room temperature. Radioactivity was counted by liquid scintillation using a 1400 Tri-Carb Packard scintillation counter. Each point was performed in triplicate and the results were expressed as mean ± SD.
Alkaline phosphatase activity assay
HT-29 cells were plated as a nondifferentiated (subconfluent) or confluent density as differentiated culture in 6-well plates. Then, the different SCFA were added at the indicated concentrations for 48 h. Cell differentiation was measured by alkaline phosphatase (AP) activity as previously described (Sanchez de Medina et al. 1996). The AP activity was determined spectrophotometrically using disodium p-nitrophenylphosphate as substrate and the results expressed as mU/mg protein.
HT-29 cells at different confluent states and subconfluent FHC cells were incubated in 24-well plates in 1 ml of complete medium in the presence or absence of the indicated SCFA. After 24 h of incubation at 37°C, the viability was assessed by crystal violet staining as previously described (Xaus et al. 2001) with minor modifications. The cells were stained and fixed with 0.2% crystal violet in 2% ethanol during 30 min at room temperature. After four washes with PBS, the cells were scrapped with 1% SDS during 30 min and then collected and centrifuged at 3,000 rpm for 5 min. Finally, the color intensity was quantified using an ELISA reader at 540 nm. Each sample was analyzed in triplicate and the results were represented as mean ± SD.
Induction of colitis and assessment of colonic damage
Female Wistar rats (170–190 g) obtained from the Laboratory Animal Service of the University of Granada (Granada, Spain) were housed individually in makrolon cages and maintained in an air-conditioned atmosphere (22°C) with a 12-h light–dark cycle, and they were provided with free access to tap water and food. All animal studies were carried out in accordance with the Directive for the Protection of Vertebrate Animals used for Experimental and other Scientific Purposes of the European Union (86/609/EEC) in compliance with the Helsinki declaration. The rats were randomly assigned to three groups (n=20): two groups (noncolitic and control groups) received a control chow diet (fiber content, 50 g/kg in the form of cellulose) and the other group (FOS group) received a FOS diet where the source of fiber was 50 g/kg diet in the form of FOS, obtained from Puleva Biotech SA (Granada, Spain). Two weeks after starting the experiment, fecal samples were obtained from rats fed with control and FOS diet and subsequently homogenized for SCFA quantification. At this point, ten rats per group were killed in order to evaluate normal healthy intestinal morphology. Then, the rest of the rats were fasted overnight, and those from the control and FOS groups were rendered colitic by the method originally described by Morris et al. (1989) with minor modifications. Briefly, they were anesthetized with halothane and given 10 mg of trinitrobenzensulphonic (TNBS) acid dissolved in 0.25 ml of 50% ethanol (v/v) by means of a Teflon (Dupont, Wilmington, Del) cannula inserted 8 cm into the anus. All rats were killed with an overdose of halothane 1 week after the induction of colitis. Once the rats were killed, the colon was removed aseptically, placed on an ice-cold plate, and longitudinally opened, and the luminal contents were also collected for SCFA determinations. Afterwards, the colonic segment was cleaned and blotted on a filter paper; each specimen was weighed, and its length was measured under a constant load (2 g). The macroscopically visible damage was scored on a 0–10 scale by two observers unaware of the treatment, according to the criteria described by Bell et al. (1995), which take into account the extent as well as the severity of colonic damage. Finally, a colon homogenate was used for PCNA western blot analysis.
Fecal homogenates and short-chain fatty acid quantification
Fecal and colonic homogenates were prepared with 150 mM NaHCO3 (pH 7.8) (100 mg/ml) in an argon atmosphere to quantify the SCFA concentration. Homogenates were incubated for 24 h at 37°C and stored at −80°C until extraction. The extraction of the SCFAs and their quantification was performed as previously described (Rodriguez-Cabezas et al. 2002). In parallel studies, fecal homogenates were centrifuged at 10,000g for 15 min at 4°C and filtered through a 22 μm syringe filter. The supernatants obtained were used as inhibitors of cell proliferation at the indicated dilutions.
PCNA Western blot analysis
Colonic samples obtained from rats were homogenized in (1/3 w/v) PBS with 0.1% SDS, 0.1% sodium deoxycholate, 1% Triton X-100, and protease inhibitors. Western blots were performed as described before (Comalada et al. 2005) with minor modifications. Cell lysates (100 μg) were loaded and membranes were incubated with PCNA antibody (1:3,000) (Transduction Laboratories, BD Biosciences, Madrid, Spain). A primary antibody against α-actin was used as a loading control and purchased from Amersham Corporation (Arlington Heights, IL, USA). Peroxidase-conjugated α-mouse IgG was used as a secondary antibody.
All in vivo results are expressed as mean ± SEM, and the in vitro results are expressed as mean ± SD. Differences between means were tested for statistical significance using a one-way analysis of variance and post hoc least significance test. All statistical analyses were carried out with the Statgraphics 5.0 software package (STSC, Md), with statistical significance set at P<0.05.
The proliferation of HT-29 is inversely correlated with cellular confluence and differentiation
When HT-29 cells were maintained for a long period of time under normal culture conditions in a confluent manner, it is well established that they display a differentiated and polarized epithelial phenotype and a reduction of the proliferation rate, similar to the normal epithelial phenotype in the colon (Chantret et al. 1988). The IEC differentiation can be indirectly measured by analyzing diverse enzymatic activities such as AP (Siavoshian et al. 1997a, b).
In a similar way, and in order to analyze if the HT-29 proliferative rate also depends on the cellular confluence, 1×105, 5×105, and 15×105 cells/ml were incubated in 24-well plates and the 3H-thymidine incorporation assay was performed (Fig. 1b). Although HT-29 are always proliferating in the presence of complete medium (DMEM + 10% FBS), the increase in the confluent state of the culture supposed a reduction of the proliferative rate of the cells (Fig. 1b). Thus, cellular confluence increases maturation and differentiation of HT-29 cells leading to a reduction in the proliferative rate of these cells.
Butyrate inhibits HT-29 proliferation, and induces the maturation and apoptosis of these cells in a confluency/differentiation-dependent way
The next objective was to determinate whether the effect of three main SCFA produced during colonic fermentation of dietary fiber on the proliferation, differentiation, and cellular viability of HT-29 colonic cancer cells also depends on cellular confluence. Since the amount of SCFA produced in normal conditions differs among them (D’Argenio and Mazzacca 1999), different physiological concentrations of each SCFA were tested: sodium butyrate (1, 2, 4, 8 mM), sodium propionate (2, 4, 8, 16 mM), and sodium acetate (4, 8, 16, 32 mM).
In a differentiated culture, and under these conditions, none of the SCFA at any of the concentrations tested inhibited the HT-29 proliferation (Fig. 2b), suggesting that cell differentiation could be involved in the protection against SCFA inhibitory effects.
Similarly to that observed on cellular proliferation, butyrate, or the other SCFA, did not modify the AP activity in confluent or differentiated HT-29 cultures (Fig. 3b). Since similar results were observed with nonconfluent and subconfluent cultures, only subconfluent cultures will be used from this point on and they have been considered as nondifferentiated cultures in contraposition to confluent differentiated HT-29 cultures.
SCFA does not affect the proliferation and viability in normal epithelial cells
The SCFA mix produced in vivo after colonic fiber fermentation inhibits HT-29 proliferation
Short-chain fatty acid (SCFA) production (μmol/100 mg feces) in the colonic fecal contents obtained from rats fed with control chow diet (50 g/kg fiber content in the form of cellulose) and FOS-supplemented diet (50 g/kg fiber content in the form of fructo-oligosaccharides) for 2 weeks
Colon length/weight ratio
SCFA mix produced by fiber fermentation do not inhibit the in vivo epithelium regeneration
Finally, we analyzed the potential effect of SCFA production in an experimental model of Inflammatory bowel disease (IBD) in rats, where the regeneration of the inflamed mucosa is a crucial point for the abrogation of the pathology. In this sense, it has been reported that dietary fiber administration to colitic rats promotes colonic butyrate production and exerts an intestinal anti-inflammatory effect in these experimental models of rat colitis (Cherbut et al. 2003; McIntyre et al. 1993). For this reason, we evaluated the production of SCFA in colonic contents of rats fed with a control or FOS-supplemented diet after the induction of a colonic inflammation, by using the TNBS model of rat colitis.
Short-chain fatty acid (SCFA) production in the colonic contents from noncolitic rats, or TNBS control colitic rats fed with control diet and TNBS colitic rats fed with a FOS-supplemented diet 1 week after colitis induction
Damage scorea (0–10)
The balance among cell growth, differentiation, and cell death maintains tissue homeostasis in the colonic crypt. Deregulation of these processes plays an important role in colonic carcinogenesis. Colorectal cancer is the second leading cause of cancer deaths in Western populations (Parker et al. 1996), and its occurrence is commonly ascribed to the transformation of normal colonic epithelium to adenomatous polyps and ultimately invasive cancer.
According to the model proposed by Vogelstein et al. (1988), cancer develops as a consequence of genetic alterations, including mutations of the p53 gene, which accumulate over one or two decades (Graz and Cowley 1997). However, numerous epidemiological studies have suggested that environmental factors strongly influence its incidence. Experimental data have linked dietary composition with colorectal carcinogenesis: a high intake of animal fat provides an increased risk while a high intake of dietary fiber provides a protection against cancer (Potter 1996). Metabolism by colonic bacterial microbiota of dietary fiber generates a high concentration (60–150 mM) of SCFAs, the most representative of that is butyrate (Hill 1995; Jacobs 1987). In fact, different in vivo studies have established the correlation between increased luminal butyrate levels and a decrease in the incidence of colon cancer (McIntyre et al. 1993; Medina et al. 1998). Moreover, it has been shown to induce growth inhibition and terminal differentiation in a variety of human colon cancer cell lines (Augeron and Laboisse 1984; Whitehead et al. 1986).
However, the mechanisms by which butyrate and other SCFA regulate cell proliferation/differentiation and apoptosis are still unclear. In this sense, it has been described that butyrate is able to block cell proliferation, mainly in the G1 phase of the cell cycle (Darzynkiewicz et al. 1981). This effect could be mediated by inhibition of the histone deacetylase activity (Whitlock et al. 1980), an increased cyclin D and p21Waf1 expression (Siavoshian et al. 1997a, b) or a decreased expression of the proto-oncogenes c-src and c-myc (Foss et al. 1989; Souleimani and Asselin 1993). In counterpart, butyrate has also been described to reduce the levels of apoptosis inhibitors such as Bcl-2 and Bc-XL together with the upregulation of proapoptotic Bak and Bax, or the induction of caspase-3 protease activity in some tumoral cell lines (Mandal et al. 2001; Ruemmele et al. 1999).
Besides the antitumoral properties of butyrate, the SCFA production and their absorption are closely related to the nourishment of colonic mucosa (D’Argenio and Mazzacca 1999). While butyrate is totally oxidized by colonocytes, the acetate and propionate could be absorbed intact and could modulate lypogenesis (Edwards 1994) and glyconeogenesis (Rombeau and Kripke 1990). For this reason butyrate has been proposed as the major source of calories used by colonocytes for their normal activity and growth.
Considering all these previous facts, several questions could be raised regarding the functions that are really exerted by butyrate and other SCFA on epithelial homeostasis. How can butyrate regulate normal epithelial proliferation while inhibiting tumoral cell growth and induces apoptosis? Do the other SCFA perform any functions regulating intestinal proliferation? What is the effect of SCFA production during the normal proliferation rate of the epithelial cells? And are they involved in the antiproliferative effects of dietary fiber thus accounting for its reported protection against colonic cancer? Finally, do they play a role in epithelial regeneration during disease recovery? In the present study, we have tried to clarify some of these discrepancies.
For this purpose, we have used the HT-29 epithelial cell line in the in vitro studies. These cells are interesting since although they are tumoral cells of intestinal origin and used for oncologic studies (Marchetti et al. 1997; Milovic et al. 2000), it has also been described that HT-29 cells could be used as in vitro models of normal epithelial cells since they can reversibly display some structural and functional features resembling those of human normal mature IEC (Andoh et al. 2001; Kalina et al. 2002). These different considerations (normal vs. tumoral cells depending on the study) could lead to confusion.
Our results showed that the proliferation/differentiation/viability characteristics of HT-29 cells depend on the confluent and phenotypic status. In this sense, we have observed that a confluent culture of HT-29 (differentiated culture), in a similar way to normal nontumoral epithelial FHC cells, showed characteristics of normal epithelial cell. This conclusion is based on the fact that although they can proliferate for the daily regeneration, they do it with a low ratio while displaying differentiation markers such as high AP activity (Siavoshian et al. 1997a, b). On the other hand, subconfluent HT-29 cultures with a high proliferation ratio and low AP activity may be characteristic of nondifferentiated colon tumoral cells (Marchetti et al. 1997). Although our results do not explain a mechanistic motive of these differences, it might explain some of the discrepancies observed in the different studies using these cells with different phenotypic states.
The present study also reveals that, among the three SCFA studied, butyrate is the most potent inhibitor of proliferation and is able to induce differentiation and apoptosis in nondifferentiated HT-29. However, propionate exhibits a weaker antiproliferative effect, while acetate is ineffective, in accordance with previous results reported by others (Milovic et al. 2000). We have not been able to demonstrate an apoptotic and differentiation effect produced by propionate in HT-29 at any of the differentiation status nor concentrations tested, in contrast to that observed by others (Gryfe et al. 1997). These discrepancies could be due to differences in the experimental protocols, since these authors measured the HT-29 differentiation state after 9 days of propionate incubation while our results are expressed only after 24–72 h.
So far, our in vitro results suggest that butyrate could exert antitumoral effect since it is able to inhibit proliferation and induce apoptosis in nondifferentiated high-proliferative intestinal cells typical of tumor progression; but at the same time, it allows normal intestinal proliferation and regeneration, since butyrate does not exert any antiproliferative effect upon normal differentiated epithelial cells. More studies are required in order to elucidate the mechanisms and the signaling pathways used by butyrate or changed during cellular differentiation in order to modify the cellular proliferation and apoptotic rate.
To confirm this hypothesis, we extended our results by performing different in vivo studies. Initially, we tested the effect of an increased SCFA production on normal intestinal architecture using rats fed with FOS as a source of dietary fiber. The mixture of SCFA produced and extracted from the feces of these animals was almost 2.5 times higher in butyrate and propionate levels than those obtained from control healthy rats. In addition, intestinal morphology, size, weight, and macroscopic and microscopic aspect were totally normal, thus suggesting a lack of any deleterious effect attributable to the higher SCFA production. In spite of that, the same mixture of SCFA was still able to inhibit the nondifferentiated HT-29 proliferation in vitro. However, although the metabolites in the fecal homogenate responsible for this inhibition are likely to be the increase on SCFA, changes of other factors caused by the consumption of fiber-rich diets, such as changes in bile acids or other metabolites resulting from microbial activity that could also be extracted from the feces, cannot be excluded.
Subsequently, we decided to test whether the nondeleterious effects of SCFA were also seen on normal intestinal growth in an animal model where an increase of intestinal proliferation is needed in order to promote the regeneration of a damaged mucosa, as it may occur in the inflammatory bowel disease (Furrie et al. 2005; Tokumasa et al. 2004).
Indeed, many reports have revealed alterations in the SCFA and organic acid concentration of the colon, especially increased lactate and decreased butyrate, in IBD patients (Araki et al. 2002). The mechanisms responsible for these alterations, however, remain unclear. It is probable that the alterations in the SCFA and organic anion levels in IBD patients may be partly due to intracellular components derived from microflora destroyed under hypotonic and aerobic conditions in the colonic lumen, for example caused by mucosal bleeding. These alterations may influence the pathogenesis and progression of IBD (Araki et al. 2002).
Our results confirm these observations since FOS supplementation to colitic rats resulted in an increased production of luminal SCFA, partially counteracting the reduction observed as a consequence of the colonic damage, and it was associated with an amelioration of the colonic inflammatory process induced by TNBS. In addition, and, confirming the benefits of the SCFA production, the increased butyrate production was not deleterious for the normal regeneration of the intestinal damaged mucosa, since increased levels of proliferation markers, such as PCNA (Waseem and Lane 1990), could be detected in spite of the increase in butyrate in the FOS-treated rats.
In conclusion, our results suggest that consumption of dietary fiber, like FOS, can enhance the production of colonic SCFA, especially butyrate, which may exert the antitumoral and anti-inflammatory properties in the colon, while maintaining normal proliferation and regeneration potential of the normal intestinal cells, and this is probably due to the phenotypical specificity of its effects.
This work was supported by grants from Spanish Ministry of Science and Technology (SAF2002-02592) and from Instituto de Salud “Carlos III” (PI021732), with funds from the European Union, and from Junta de Andalucía (CTS-164). EB is recipient of a fellowship from the Ministry of Science and Technology, Spain. MC is member of the “Juan de la Cierva” program, Ministry of Science and Technology, Spain. The authors want to thank Dr. M.E. Rodríguez-Cabezas for editorial assistance and Dr. M.N. Rodríguez-Cabezas for technical assistance.