Chronic, severe protein malnutrition results in a marked reduction in whole body protein turnover in most mammals (Golden et al. 1977). However, the extent to which decreased whole body protein turnover reflects decreased protein turnover in individual organs has never been clear, particularly with respect to the small intestine (Garlick et al. 1975, Hirschfield and Kern 1969, McNurlan and Garlick 1981, McNurlan et al. 1982, Waterlow and Stephen 1968). Recently in an attempt to clarify the effects of protein malnutrition on individual organs, we examined the adaptation of the small intestine of young pigs to dietary protein deficiency and demonstrated a marked decline in the rate of total protein synthesis in protein-malnourished compared with well-nourished animals (Wykes et al. 1996). However, the intestinal mucosa contains numerous cell types, each with specific functions. Thus, the implications of the observed changes in total intestinal protein turnover from the functional viewpoint were difficult to infer.

We have now extended our investigation to examine the effect of chronic protein deprivation on the synthesis of a specific small intestinal protein which is of key functional importance for young animals, brush border lactase phlorizin hydrolase (BB LPH).⁴ LPH is an essential digestive enzyme for all suckling mammals because it hydrolyses lactose, the predominant sugar of mammalian milk, to galactose and glucose that can be absorbed. The synthesis and abundance of this enzyme, therefore, is of critical importance to the suckling mammal.

BB LPH is a glycoprotein synthesized only in the villus enterocyte. Enzyme synthesis is a complex process controlled by a series of transcriptional and post-transcriptional events that culminate in insertion of the mature protein into the BB membrane (Dudley et al. 1992b, 1993 and 1994a). Gene transcription starts in the cells at the base of the villus (Dudley et al. 1992b). The first detectable precursor form of the enzyme is translated and cotranslationally glycosylated in the endoplasmic reticulum to form a high-mannose LPH precursor (proLPH_h, Dudley et al. 1996). In the Golgi apparatus, proLPH_h is converted to the complex glycosylated precursor (proLPH_c) which is in turn translocated to the BB membrane (Dudley et al. 1996). Either during translocation or immediately following insertion of the enzyme into the BB membrane, proLPH_c is proteolytically cleaved to form the mature BB protein (Dudley et al. 1996). Thus, as a result of its complex biosynthesis, LPH expression in the BB membrane may potentially be controlled by factors that regulate: 1) mRNA abundance, 2) the rate of mRNA translation, and 3) the rate of multiple steps of post-translational synthesis. This study further defines the effect of protein malnutrition on the small intestine by describing the effects of protein deficiency on LPH synthesis.

MATERIALS AND METHODS

Eight 2-wk-old piglets were obtained from a herd of commercial pigs at Texas A&M University, College Station, TX. They were housed in individual pens and fed Soweena Litter Life (Merrick, Madison, WI) during a 5-d adaption period. They were then divided by weight into two experimental treatment groups and fed for 8 wk semipurified dietsa protein-deficient diet containing 30 g protein and 16.77 MJ metabolizable energy (ME) per kg diet, or a protein-sufficient diet containing 193 g protein and 15.52 MJ-ME per kg diet.

The diets were prepared as previously described (Wykes et al. 1996), and compositions are shown in Table 1. The magnesium concentration in both diets was above the recommended level of 400 mg/kg diet (NRC 1988). It should be noted that the potassium concentration in the low-protein diet was 1.7 g/kg. This value is low for very young, well-nourished, growing pigs that require 2.8 g K/kg diet (NRC 1988). However, it is above the concentration (<1.5 g/kg diet) required by older, slowly growing pigs whose growth rate is comparble to these protein-malnourished animals (NRC 1988). The potassium requirement for malnourished pigs has not been studied, but, importantly, the effect of potassium deficiency only becomes truly noticeable in very young, growing pigs when the dietary concentration of potassium is below 1.5 g/kg (Jensen et al. 1961).

Table 1. Composition of diets [View Table]

The pigs had free access to food and water and were weighed biweekly. Venous blood samples were obtained weekly by subclavian puncture for biochemical analyses (Wykes et al. 1996). After the pigs had been fed their respective diets for 8 wk and after overnight food deprivation, they were anesthetized with 5% isoflurane (Aerrane, Anaquest, Liberty Corner, NJ), and catheters were implanted in the jugular vein and carotid artery as previously described (Dudley et al. 1994a). Following surgery the pigs were monitored closely for 5 h when free access to food and water was resumed.

Three days later, each pig was infused via the jugular vein with a sterile solution of ²H₃-leucine prepared in saline (9 g/L). The well-nourished pigs were given a priming dose of 30 µmol·kg,¹ followed by a constant infusion of 30 µmol·kg¹·h¹ for 8 h. To compensate for the possibility of slower tissue protein synthesis in the protein-malnourished pigs, they received a 45 µmol·kg¹ priming dose and a constant infusion of the isotope at 45 µmol·kg¹·h¹ for 8 h.

Pigs consumed 1/12 of daily food intake offered at hourly intervals during the infusion. A 5-mL sample of arterial blood was taken from the carotid catheter before the infusion and hourly thereafter. Blood was drawn into prechilled tubes containing 50 µL of a solution of Na₂EDTA (100 g/L), sodium azide (20 g/L), merthiolate (10 g/L), and soybean trypsin inhibitor (20 g/L). Blood samples were centrifuged immediately at 4°C, and plasma was removed and stored at 70°C for later analysis (Wykes et al. 1996).

At the end of the infusion the pigs were killed by intravenous injection of 0.33 mL of Beuthenasia-D/kg (Schering-Plough Animal Health Corp, Kenilworth, NJ), and the entire small intestine, from the peritoneal reflection (analogous to the ligament of Treitz) to the ileal-cecal junction, was quickly excised and chilled in iced saline (9 g/L). The small intestine was divided into two segments of equal length, and the proximal segment was defined as the jejunum. Weighed midjejunal samples were taken and frozen in liquid N₂ for mRNA analysis. An additional midjejunal sample was weighed and fixed in phosphate-buffered formalin for histology. The remainder of the jejunum was flushed with cold saline and weighed. The mucosa was scraped and weighed, and a sample was homogenized in phosphate buffer containing protease inhibitors as previously described (Dudley et al. 1992a). Aliquots of the homogenate were frozen at 70°C for enzyme and protein measurements, while the remaining homogenate was used for the immunoisolation of LPH proteins.

1

1

RNA was isolated by the guanidine isothiocyanate-cesium chloride method and was fractionated and blotted as previously described (Shulman et al. 1992). In brief, total cellular RNA (10 µg per lane) was fractionated on agarose gels (8 g/L) containing 2.2 mol formaldehyde/L in MOPS buffer (0.02 mol 3-(N-morpholino)propanesulfonic acid/L). Ethidium bromide staining of the RNA was used to assess the integrity of the RNA. All data are derived from lanes in which staining of the ribosomal RNA bands indicated that the RNA was intact and that the lanes were evenly loaded. Electroblotting for 3-4 h was used to transfer the RNA to a nylon membrane. RNA was UV cross-linked to the membrane.

Radioactive probes were prepared by random-primed labeling of the linearized plasmids (Chandrasena et al. 1992, Shulman et al. 1992). Blots were prehybridized at 42°C for approximately 4 h in a solution of 50 mL deionized formamide/100 mL, 3× SSC, 5× Denhardt's solution, 0.05 mol NaPO₄/L (pH 7.4), 1 g sodium dodecylsulfate/L (SDS), and 250 µg salmon sperm DNA. Hybridization (16-20 h at 42°C) was performed in the same buffer with 100 g dextran sulfate/L and 40 MBq/L of the ³²P-labeled probe. Blots were washed and hybridization detected as previously described (Chandrasena et al. 1992, Shulman et al. 1992). After autoradiography to visualize the positions of the bands, the relative abundance of LPH mRNA was quantified by cutting the bands representing LPH or EF-1 mRNA from the blot and measuring their radioactivity in a liquid scintillation counter. The abundance (Bq) of LPH mRNA was calculated relative to the abundance (Bq) of EF-1 mRNA.

The use of the equation involves two assumptions: first, that, as we have shown in rats and pigs (Dudley et al. 1993 and 1996), isotopic equilibrium is achieved rapidly in both the mucosal amino acid pool and proLPH_h. Second, that once isotopic equilibrium had been achieved in proLPH_h, label incorporation into the BB protein is linear. The validity of this assumption we previously demonstrated (Dudley et al. 1992a, 1993 and 1996). However, we have also shown a delay of ~1 h as label moves from proLPH_h to proLPH_c and finally to the mature BB protein. Thus, for both experimental groups, because we estimated the rate of BB LPH synthesis from 0 h, these calculations slightly underestimate the true rate of label incorporation into the BB protein.

1

1

Table 2. Brush border lactose activities, mRNA ratios, and polypeptide relative amounts in midjejunum of well-nourished and protein-malnourished piglets1,2 [View Table]

The calculation of total abundance is based on the assumption that lactase activity is attributable only to the mature BB form of the enzyme. Naim et al. (1991) have reported that in COS-1 cells transfected with the human cDNA for LPH, the precursor forms of the enzyme appear to be enzymatically active. However, in these cells, the mature BB (160 kDa) form of the enzyme, seen in vivo, is not synthesized, and a precursor form of LPH appears to be expressed at the cell surface instead. Thus, until it has been conclusively proven that the precursor polypeptides in the pig are enzymatically active, it seems reasonable to omit these proteins from all calculations of total abundance. More importantly for this study, the small difference in the relative amount of BB LPH between the two experimental treatment groups (<2%) is not sufficient to alter substantially the calculations of BB LPH absolute synthesis rates.

RESULTS

1

Plasma albumin concentrations among pigs were not different (average 31 ± 2 g·L¹) at the start of the study (Wykes et al. 1996). After 6 wk, plasma albumin concentrations had declined significantly in the protein-malnourished pigs to 48% of the prediet value while they had increased by 21% in the well-nourished pigs (Wykes et al. 1996).

The jejunal weight was significantly (P < 0.05) higher in the well-nourished than in the protein-malnourished pigs (331 ± 9 g and 125 ± 15 g, respectively) but jejunal weight·kg body weight¹ was not significantly different (0.97 ± 0.05 and 1.1 ± 0.09 g/100 g for well-nourished and protein-malnourished pigs, respectively). The weight of the scraped mucosa was also significantly (P < 0.05) greater in the well-nourished than in the protein-malnourished pigs (240 ± 13 g and 90 ± 8 g, respectively), but was a constant proportion of jejunal weight regardless of experimental treatment (mean of all pigs 72 ± 3 g/100 g). Protein per gram of scraped mucosa was significantly (P < 0.05) higher in the well-nourished than in the protein-malnourished pigs (159 ± 5 mg/g and 130 ± 5 mg/g, respectively). In contrast, the RNA concentrations were not different (1.3 ± 0.08 mg/g mucosa in well-nourished and 1.2 ± 0.12 mg/g mucosa in protein-malnourished pigs). Lactase activities did not differ between groups (Table 2).

Villus height and crypt depth were significantly (P < 0.05) greater in the well-nourished than in the protein-malnourished pigs. In well-nourished pigs the villus height was 340 ± 24 µm and the crypt depth was 252 ± 38 µm, while in protein-malnourished pigs the villus height was 247 ± 7 µm and crypt depth was 192 ± 6 µm; Figure 2a and b). The jejunal enterocytes appeared to be morphologically normal in both groups. The cells were columnar with nuclei located close to the basolateral membrane. There were 55 ± 4 cells per 500 µm midvillus segment regardless of experimental treatment. However, there appeared to be more cells in the lamina propria of well-nourished than protein-malnourished pigs (Figure 2c and d).

Figure 3 illustrates a representative northern blot of LPH mRNA and EF-1 mRNA in each pig. When the EF-1 bands were counted, no significant differences between protein-malnourished and well-nourished pigs (29 ± 2 Bq and 27 ± 2 Bq, respectively) were observed. In contrast, LPH mRNA abundance was noticeably lower in the protein-malnourished pigs than in well-nourished pigs. As a result, LPH mRNA abundance relative to EF-1 mRNA abundance was significantly (P < 0.05) higher in well-nourished than in protein-malnourished pigs (Table 2).

Following SDS-PAGE, LPH polypeptides immunoisolated from the solubilized, scraped mucosa, separated into 4 bands (Figure 1), which we have previously identified as two precursor forms of BB LPH (proLPH_h, 200 kDa and proLPH_c, 220 kDa) and two forms of BB LPHthe 160-kDa polypeptide and a dimer of BB LPH with an apparent molecular mass of approximately 240 kDa (Dudley et al. 1993, 1994a and 1996). In order of synthesis, the LPH polypeptides are proLPH_h (the first detectable translation product), proLPH_c (the complex glycosylated precursor) and, finally, the mature BB enzyme (Dudley et al. 1994a and 1996).

The relative amounts (i.e., the amount of an individual LPH polypeptide relative to the total amount of all four LPH polypeptides on a gel) appeared different between groups (Table 2), however, the difference was significant (P < 0.05) only for proLPH_c.

The enrichment of the plasma free amino acid pool reached steady state within an hour (Wykes et al. 1996). After 8 h, the enrichment of the mucosal free amino acid pool was 4.3 ± 0.4 mol/100 mol for the well-nourished pigs and 11.6 ± 0.9 mol/100 mol in the protein-malnourished pigs. In both groups, however, the enrichment of the mucosal free amino acid pool was a constant percentage of the plasma free pool (mean for all pigs: 27 ± 2%).

The enrichment of the earliest detectable precursor polypeptide (proLPH_h) was 4.2 ± 0.3 mol/100 mol for the well-nourished pigs and 16.2 ± 1.5 mol/100 mol for the protein-malnourished pigs. The enrichment of proLPH_h as a percentage of the enrichment of mucosal free amino acid pool was significantly (P < 0.05) different between groups (100 ± 11% for well-nourished pigs vs. 140 ± 7% for protein-malnourished pigs).

The enrichment of proLPH_c was 4.1 ± 0.6 mol/100 mol for the well-nourished pigs and 13.4 ± 1.5 mol/100 mol for protein-malnourished pigs. The enrichment of proLPH_c as a percentage of the enrichment of proLPH_h did not differ between groups (95 ± 9% vs. 82 ± 3%, well-nourished vs. protein-malnourished pigs). BB LPH did not reach the enrichment of proLPH_c during the 8 h infusion (1.46 ± 0.2 mol/100 mol in well-nourished pigs and 3.5 ± 0.3 mol/100 mol in protein-malnourished animals). The enrichment of the dimer of BB LPH (240 kDa polypeptide) paralleled that of the 160 kDa BB protein (data not shown).

The total abundance (arbitrary units) of BB LPH per gram of tissue was not significantly different between the two treatment groups, but the K_s of BB LPH synthesis was significantly (P < 0.05) higher in the well-nourished than in the protein-malnourished pigs (Table 3). Thus, the absolute synthesis rate (a measure of the amount of LPH synthesized per gram of tissue) was, likewise, significantly (P < 0.05) higher in the well-nourished than in the protein-malnourished pigs (Table 3).

Table 3. Total brush border lactase phlorizin hydrolase (BB LPH) abundance, the rate constants (Ks) of BB LPH synthesis and absolute synthesis rates of BB LPH in well-nourished and protein-malnourished piglets1 [View Table]

DISCUSSION

LPH mRNA abundance relative to EF-1 mRNA abundance was significantly lower in the protein-malnourished than in well-nourished pigs. These findings are consistent with studies in the rat showing that LPH mRNA abundance can be modified by diet (Hodin et al. 1994). In rats, for example, the enterocyte responds to food intake following food deprivation by increasing the expression of alkaline phosphatase mRNA and reducing the expression of LPH mRNA (Hodin et al. 1994). Further, while LPH mRNA expression in rats normally decreases at weaning, the disappearance of the transcript can be delayed by prolonged nursing (Dudley et al. 1992b, Duluc et al. 1992 and 1993).

The lower LPH mRNA levels in protein-malnourished compared to well-nourished pigs may relate to the intestinal morphology of the dietary treatment groups. Villus height in the well-nourished pigs was approximately twice that in the protein-malnourished pigs. Since the number of cells in a 500 µm mid-villus segment was not affected by dietary treatment, enterocyte size apparently did not differ between treatment groups. It seems likely, therefore, that there were fewer villus enterocytes in protein-malnourished than in well-nourished pigs, and, because the villus enterocyte is the only cell capable of LPH mRNA expression, the abundance of LPH mRNA was necessarily reduced (Dudley et al. 1992b, Keller et al. 1992).

Dietary treatment also affected the steps of post-translational LPH synthesis. The relationship between the steady-state enrichment of proLPH_h and the intracellular free amino acid pool, for example, differed significantly between groups. Since we have shown in rats and pigs that proLPH_h reaches isotopic equilibrium within 60 min using the present isotope infusion strategy (Dudley et al. 1993, 1994a and 1996), it seems reasonable to assume that after 8 h of infusion, proLPH_h was at isotopic steady state. However, proLPH_h steady-state enrichment as a percentage of the enrichment of the mucosal free amino acid pool was significantly higher in the protein-malnourished compared to the well-nourished pigs (140% vs. 100%, respectively). These findings suggest that the mucosal free pool from which proLPH_h is synthesized is compartmentalized, and that different compartments may be used for proLPH_h synthesis in well-nourished and protein-malnourished pigs. We have previously demonstrated similar findings for sucrase-isomaltase (SI) synthesis in fed and food-deprived rats (Dudley et al. 1992a). In the latter case, the steady-state enrichment of pro-SI_h was 148% of the mucosal free amino acid pool in fed rats and 241% in food-deprived rats.

The K_s and the absolute synthesis rate (in arbitrary units, K_s multiplied by total abundance) for BB LPH were significantly lower in protein-malnourished than in well-nourished pigs. Again, these findings are consistant with earlier studies of the mechanisms regulating hydrolase expression (Dudley et al. 1992a, Keller et al. 1992, Olsen et al. 1986). In pig intestinal cultures, for example, dietary fructose suppressed the posttranslational synthesis of both SI and aminopeptidase N (Danielsen 1989). In postweaned rats, increasing dietary carbohydrate increases the in vivo absolute synthesis rate of BB LPH (Goda et al. 1995). We have shown that the rate constant of SI synthesis in adult rats is determined by whether the rats are fed or food deprived (Dudley et al. 1992a) and have recently demonstrated that the in vivo steps of post-translational synthesis of LPH can be modified in neonatal pigs by feeding colostrum (Burrin et al. 1994, Dudley et al. 1996). It is important to point out, however, that in this study the calculated in vivo synthesis rates for BB LPH are likely to be lower than the actual rates. As discussed above, for this study we wished to examine LPH synthesis in conscious feeding pigs and were, therefore, able to obtain small intestinal tissue samples only at the end of the 8-h infusion. Because of the delay (see Calculations) before isotope reaches the BB LPH compartment, the enrichment of tissue samples obtained at the end of the infusion may not accurately reflect the rate of BB LPH enrichment. Under these conditions, the K_s is an underestimate of the actual synthesis rate. However, since this is true for both dietary treatment groups, comparison between groups accurately reflects the effect of protein-malnutrition on BB LPH synthesis.

To summarize, this study demonstrates that in protein-malnourished pigs fed high-carbohydrate diets, LPH synthesis is regulated by mRNA abundance and the rate of post-translational synthesis. Steady state mRNA abundance and the K_s for BB LPH synthesis from proLPH_h in protein-malnourished were on average ~58% and ~65%, respectively (see Tables 2 and 3), of the values in well-nourished pigs. The data do not, however, define the predominant mechanism whereby LPH expression is controlled in these protein-deficient pigs. Interestingly, a comparison of the absolute synthesis rate/unit of LPH mRNA (LPH mRNA/EF-1 mRNA) in the two groups reveals that the amount of LPH synthesized/unit of LPH mRNA did not differ (2532 ± 386 for well-nourished pigs and 2577 ± 430 for protein-malnourished pigs). While these calculations were not based on the direct measurement of mRNA abundance (i.e., by a RNase protection assay) and are undoubtedly imprecise, this observation suggests that the expression of BB LPH is controlled chiefly by LPH mRNA abundance. If post-translational synthesis had been the major regulatory mechanism, one would expect the absolute synthesis rate/unit LPH to be markedly lower in protein-malnourished than in well-nourished pigs.

ACKNOWLEDGMENTS