Both L- and D-lactate contribute to metabolic acidosis in diarrheic calves.

Diarrhea in neonates is often complicated by metabolic acidosis. We used blood gas analysis and HPLC to determine whether bacterial fermentation might contribute to acidosis in diarrheic calves. Diarrheic calves (n = 21) had significantly lower pH, PCO(2), HCO(3)(-) and a higher anion gap than healthy calves (n = 21). Serum concentrations (mean +/- SD, mmol/L) of DL-, L- and D-lactate were also significantly higher in diarrheic (8.9 +/- 5.1, 4.1 +/- 3.4 and 5.2 +/- 5.7) than in healthy calves (1.7 +/- 1.2, 2.0 +/- 1.1 and too low to quantify). D- and L-lactate accounted for 64% anion gap increase in diarrheic calves. Fecal D- and L-lactate concentrations were also significantly higher in diarrheic calves (9.4 +/- 3.0 and 11.9 +/- 2.7 mmol/L) than healthy calves (1.1 +/- 0.1 and 1.6 +/- 0.1 mmol/L). The elevated concentrations of serum and fecal D-lactate suggest gut bacterial fermentation contributes to the development of acidosis in diarrhea.

Neonatal diarrhea is a major health concern in human infants (1) and in many other species, including calves (2). Metabolic acidosis is an important complication of diarrhea regardless of the causal agents (3, 4). Metabolic acidosis may contribute to cardiac arrhythmia or death from heart failure by decreasing myocardial potassium, elevating extracellular potassium and changing membrane potential (5, 6). Therapeutic studies show that correction of acidosis is essential to the well being of calves (7) and humans (4) with diarrhea. Furthermore, preventing the development of acidosis has long-term benefits on the demeanor and survival of diarrheic calves (8). Because acidosis has such an important bearing on the severity of illness and outcome in diarrhea, a proper understanding of its origins and nature is important.

The most commonly accepted causes of metabolic acidosis in diarrheic humans and calves are fecal bicarbonate loss and L-lactic acidosis (9, 10). L-lactic acidosis is thought to arise from poor tissue perfusion because of dehydration or endotoxemia with subsequent anerobic glycolysis and decreased hepatic clearance of L-lactate (11, 12, 13, 14, 15, 16). We postulate that L-lactate is unlikely to be the only organic acid responsible for acidosis in diarrheic calves because in a study of neonatal diarrheic calves anion gap was increased in 24 of 53 calves but only 3 calves with high anion gaps had increased plasma L-lactate (3). Although D-lactic acidosis is not presently regarded as a complication of enteritis, it has been recorded as a complication of jejunoileal by-pass, short bowel syndrome (SBS)3 in humans (17, 18) or metabolic acidosis without dehydration in calves with no or minimal diarrhea (19). We propose that D-lactic acidosis might contribute to metabolic acidosis in diarrheic calves through mechanisms similar to those that operate in jejunoileal by-pass.

Because diarrhea causes malnutrition (20), several bodies, including the World Health Organization (1), recommend the continuation of appropriate food administration during diarrhea against the traditional practice of withholding food. This preserves the patient's nutritional status (21). However, gastrointestinal dysfunction resulting in diarrhea may be accompanied by small intestinal malabsorption (22). Unabsorbed nutrients may subsequently be fermented by colonic bacteria, as occurs in patients with SBS (18). Gut fermentation may lead to the production of lactate, in which case both the D and L isomers should be produced (23). Because the L-form but not the D-form is rapidly metabolized (24, 25) in the body, the D-form may accumulate in the blood and be a contributor to acidosis in diarrheic patients.

We hypothesized that intestinal bacterial substrate fermentation contributes to acidosis in diarrhea caused by enteritis. We used a neonatal calf model because calves are fed milk, like human infants, often develop a severe acidosis when diarrheic and most causes of calf diarrhea, including enterotoxigenic Escherichia coli, rotavirus and coronavirus (26, 27) are accompanied by villous atrophy. We used stereospecific and nonstereospecific HPLC methods previously developed and validated in our laboratory (28) to detect unusual organic acids.

MATERIALS AND METHODS

Animals.

Two groups of single-suckle beef calves were used, healthy and diarrheic; all were between 1 and 45 d old. Diarrheic calves (n = 21) were selected in the study based on the finding of diarrhea on physical examination of sick calves presented to the large animal clinic of the Western College of Veterinary Medicine, University of Saskatchewan. All diarrheic calves had access to their dam's milk before presentation to the clinic by their owners. Some had been treated with oral electrolyte solutions containing glucose, glycine and electrolytes. Twenty-one calves from a local beef farm that were normal on physical examination were used as healthy controls. Healthy calves also had access to their dam's milk. The physical condition of each scouring calf was assessed at time of entry into the study (29). The study was approved by the Animal Care Committee, University of Saskatchewan and carried out in accordance with the principles and guidelines specified by the Canadian Council on Animal Care.

Experimental protocol.

At admission, blood (2.5 mL) was collected anerobically from the jugular vein of the calves in a preheparinized plastic syringe (Smooth; Radiometer America, Westlake, OH) for blood gas measurements. A second blood sample (2.5 mL) for the determination of organic acid concentration was also collected at admission and placed in a tube containing no anticoagulant (Vacutainer; Becton Dickinson, Rutherford, NJ). This blood was allowed to stand for between 20 and 40 min. On clotting, serum was separated by centrifugation and frozen at −20°C until analyzed. All calves were resting when the blood samples were drawn. Fecal samples were collected simultaneously after perineal massage and frozen until HPLC assay.

Blood gas measurements.

All acid-base parameters (pH, bicarbonate concentration, anion gap, partial pressure of carbon dioxide and base excess) were determined within 15 min of blood collection using an automated blood gas analyzer (Ciba Corning 288 blood gas system; Ciba Corning Canada, Markham, Ontario). Bicarbonate, pH and PCO2 values were corrected to the calf's rectal temperature. Plasma sodium, potassium and chloride concentrations were also determined simultaneously using a spectrophotometric autoanalyzer (Abbott Spectrum System, North Chicago, IL). Anion gap was measured as the difference between the sum of the serum concentrations of the readily measured cations, Na+ + K+, and the readily measured anions, Cl− + HCO3−.

Organic acid measurements.

Serum and fecal samples collected were analyzed for organic acid concentrations by HPLC 1 wk after the last sampling. Before HPLC analysis, serum samples were deproteinized by ultrafiltration. To an Ultrafree-MC centrifugal filter unit (Millipore, Milford, MA), 100 μL of standard adult bovine serum was added along with 50 μL internal standard solution and made up to 200 μL with aqueous solution of standard DL-lactate for calibration purposes. The internal standards used were citric acid and malonic acid for the nonstereospecific and stereospecific assays, respectively. For quantification purposes, aliquots (100 μL each) of each sample were used instead of standard bovine serum and 50 μL of double distilled water (DDW) replaced standard DL-lactate. The resulting solution was spun at 10,600 × g for 30 min. Aliquots of 20 μL from the filtrate were then injected into the HPLC system. For the stereospecific analysis of lactate, fecal samples (0.5 g each) were mixed with 5 mL DDW and centrifuged at 10,600 × g for 20 min. The supernatant was drawn off with a syringe and was further purified by filtration through an Acrodisc PF (0.8/0.2)-μm syringe filter (Pall Corporation, Ann Arbor, MI). These filtrates (100 μL) were subsequently deproteinized in the same manner as the serum samples, using DDW instead of standard bovine serum. Aliquots of 20 μL from the filtrates were then injected into the HPLC system.

Both the stereospecific and nonstereospecific assays made use of the same HPLC system, under different conditions as previously described (28). The HPLC system consisted of a Waters Model 600 pump, a Waters 486 tunable wavelength ultraviolet absorbance detector and a Waters 710 Ultra WISP autoinjector (Waters, Mississauga, ON). Data collection, integration and calibration were performed using a Waters Millennium chromatography manager, Version 2.1 (Waters, Mississauga, ON). The nonstereospecific analysis used a reverse-phase, 300- × 8.0-mm analytical column (Shodex RSPAK KC-811; Showa Denko K.K., Tokyo, Japan) for the chromatographic analyses of lactic acid, pyruvate, acetate and citric acid (internal standard). Lactic acid, pyruvic acid and acetic acid were identified based on their retention times with known standards of each acid and recovery was determine from the internal standard. Concentrations of pyruvate, acetate and lactate were determined from standard curves made with solutions of the appropriate acid (28). The mobile phase, consisting of 0.1% phosphoric acid and DDW, was pumped at 0.7 mL/min in the isocratic mode, with the column temperature maintained at 50°C. Ultraviolet light detection was at 205 nm, representing the ultraviolet maxima of a solution of lactate in the nonstereospecific mobile phase (28).

The stereospecific analysis of lactate enantiomers used a stainless steel 3 μm octodecylsilane packed analytical column (50 × 4.6 mm internal diameter), coated with N,N-dioctyl-L-alanine as chiral selector (Chiral Technologies, Exton, PA). The mobile phase consisted of 2 mmol/L copper sulfate containing 1% acetonitrile and was pumped in the isocratic mode at 0.4 mL/min at room temperature. Ultraviolet light detection was at 236 nm. This wavelength represented the ultraviolet maxima of a solution of lactic acid in the stereospecific mobile phase (28). The mobile phases of both assays were filtered through 0.45-μm membrane filters (Scheicher & Schuell, Keene, NH) and degassed under vacuum to remove oxygen and contaminants before use. For each assay, samples were randomly analyzed within each batch.

Statistical analysis.

Statistical data were analyzed with Analysis ToolPak program on Microsoft Excel, Version 5.0, 1985–1994 (Redmond, WA). The mean concentrations of all organic acids were calculated for the healthy and diarrheic calves. Unpaired two-sample t test assuming unequal variances was used to compare blood gas components and measured concentrations lactic acid, its enantiomers and other organic acids were measured and compared between the two groups of calves. All values were expressed as the mean ± SD, and a P value < 0.05 was considered significant.

RESULTS

Age, physical assessment and blood gases.

Average age of calves was 18.8 ± 16.4 and 16.4 ± 13.7 d in healthy and diarrheic calves, respectively. The age of calves ranged between 1 and 34 d in the healthy group and 5 and 45 d in the diarrheic group. Diarrheic calves were depressed at presentation in 19 instances, with 14 of the depressed calves in sternal or lateral recumbency. Blood gas analysis showed that healthy calves had significantly higher mean values of pH, PCO2, HCO3− and base excess and a significantly lower anion gap than their diarrheic counterparts (Table 1). There were no significant differences in means for plasma sodium, potassium and chloride concentrations between healthy and diarrheic calves.

TABLE 1

Blood gas, electrolytes and serum organic acid concentrations in healthy and diarrheic neonatal calves 1

Blood variable	Healthy	Diarrheic
	mmol/L
pH	7.36 ± 0.03	7.17 ± 0.14**
PCO2	57.8 ± 4.3	46.5 ± 12.3**
HCO3−	32.6 ± 2.1	18.0 ± 8.8**
Base excess	6.9 ± 2.3	−10.3 ± 10.0**
Na+	135.6 ± 3.1	138.1 ± 12.2
K+	5.0 ± 0.4	5.4 ± 2.1
Cl−	102.4 ± 2.0	105.8 ± 10.9
Anion gap	5.6 ± 3.5	19.7 ± 8.6**
Pyruvic acid	0.058 ± 0.034	0.107 ± 0.066**
DL-lactic acid	1.7 ± 1.2	8.9 ± 5.1**
L-lactic acid	2.0 ± 1.1	4.1 ± 3.4**
D-lactic acid	NQ (<0.5)	5.2 ± 5.7**
Acetic acid	0.2 ± 0.4	0.9 ± 1.6*

Values are means ± SD. Asterisk denotes significantly different blood parameters between healthy (n = 21) and diarrheic calves (n = 21), P < 0.05 (Student's t test). Double asterisk denotes P < 0.01. NQ indicates not quantifiable; HCO3−, bicarbonate; Na+, sodium ion; K+, potassium ion; Cl−, chloride ion; anion gap, Na+ + K+ − HCO3− − Cl−; PCO2, partial pressure of carbon dioxide.

Serum concentrations of organic acids.

Diarrheic calves had significantly higher (P < 0.05) concentrations of acetic, pyruvic, DL-,L- and D-lactic acids than healthy calves (Table 1). To determine whether age-related differences affected D-lactic acid concentrations in our diarrheic calves, the mean age of diarrheic calves with either higher D-lactate or higher L-lactate concentrations was compared. Diarrheic calves in which D-lactic acidosis predominated were 15.9 ± 6.4 d old, whereas those in which L-lactic acidosis predominated were 16.5 ± 11.0 d old.

Fecal concentrations of organic acids.

Overall, fecal samples were collected from 18 diarrheic calves and 20 healthy calves. Fecal samples were not obtained from two diarrheic calves. Fecal lactate concentrations (D and L, respectively) were significantly higher (P < 0.05) in diarrheic calves (9.4 ± 3.0 and 11.9 ± 2.7 mmol/L) compared with healthy calves (1.1 ± 0.1 and 1.6 ± 0.1 mmol/L).

DISCUSSION

This study represents the first attempt to partition the cause of metabolic acidosis in neonatal diarrheic calves between a broad spectrum of organic acids. It is also the first description of D-lactic acidosis in patients for which the main presenting complaint is diarrhea of presumed infectious origin. It is also difficult to find reports of the contribution of other intestinally derived organic acids to metabolic acidosis in patients with enteritis. Previous studies in diarrheic calves investigated the association of acidosis with L-lactate only (15, 30) and provided evidence that this was derived from tissue metabolism secondary to dehydration. The use of fast and simple HPLC methods of analysis in this study, however, afforded the simultaneous investigation of other organic acids that may contribute to metabolic acidosis.

When diarrheic calves were compared with their healthy counterparts, they had a moderately severe metabolic acidosis with respiratory compensation. This is consistent with previous studies (15, 29). The mean anion gap was increased by > 14 mmol/L in diarrheic calves. This represents the total amount of organic acids that could have been added to serum along with more minor contributions associated with changes in concentrations or charge of calcium, magnesium, phosphates and albumin (31). The concentrations of all acids measured in this study, pyruvic, lactic and acetic acids, were significantly higher in the diarrheic calves. This indicates that all acids contribute to the lower mean pH value and higher anion gap in the diarrheic calves.

Although acetic acid and pyruvic acid concentrations were higher in diarrheic calves, they were present in very low, and sometimes unmeasurable concentrations, and could only be a minor contributor to acidosis. Disposal of these acids through the tricarboxylic acid cycle could be limited by low tissue perfusion and reduced tissue oxygen supply in diarrhea. Moreover, increased production of volatile fatty acids from bacterial fermentation in the gastrointestinal tract may also result in the production of acetic acid (32).

The levels of DL-lactate in our healthy calves are consistent with typical values for blood lactate concentrations in normal ruminants, which vary between 0.5 and 2.0 mmol/L (24). In our study, concentrations of DL-lactic acid measured using the nonstereospecific assay were approximately the same as the sum of D- and L-lactic acids measured with the stereospecific HPLC method for each calf. This further demonstrates the reliability of our analytical approach.

Mean concentrations of L-lactate were elevated in diarrheic calves, similar to previous findings (15). L-lactate is normally formed by mammalian cells during anaerobic respiration (14). Its concentration is an index of tissue oxygen deprivation and is helpful in grading the severity of shock (33). It can also be produced in large amounts in exercising, healthy, muscle (34). It has been reported in neonatal diarrheic calves as a sequela to diarrhea induced dehydration (15). An anaerobic respiration component of L-lactic acidosis due to limited supply of oxygen because hypovolemia probably existed in our study. In addition, the finding of high concentrations of L-lactate in the feces suggests that some may have been of enteric origin. L-lactic acid alone, however, can only explain a small part of the acidosis in our diarrheic calves since the increase in anion gap in diarrheic versus healthy calves was in excess of 14 mmol/L, whereas the combined increases in L-lactate and pyruvate was < 3 mmol/L. Others have reported similar findings (3).

Serum D-lactate concentrations were higher in diarrheic calves, indicating that D-lactic acidosis is a contributory factor in calves with high anion gap acidosis. D-lactic acidosis was more severe than lactic acidosis in nine calves. The ages of the high L- and D-lactic acidosis groups were similar, so age-related differences in body metabolism or intestinal flora were not responsible for the difference in types of lactic acidosis in this study. Overall, increases in D- and L-lactate totaled ∼9 mmol/L and, thus, together accounted for 9 mmol/L of the 14-mmol/L increase in anion gap.

Studies have previously linked D-lactic acid production with bacterial fermentation in the gastrointestinal tract (18, 35, 36). Malabsorption of nutrients in the small intestine of SBS patients, with subsequent fermentation by colonic bacteria results in D-lactic acidosis in humans (18, 35, 36, 37). Malabsorption and passage of undigested nutrients into the distal small intestine may cause the same process in diarrheic calves. Previous studies of diarrheic calves at our college show that many are infected by organisms known to cause villus atrophy (38) and, consequently, maldigestion and malabsorption of nutrients (39). As a result, excessive delivery of nutrients to the distal small intestine and large colon could result in bacterial overgrowth and probably D-lactic acid production. Presumably, milk lactose could be readily fermented to D- and L-lactic acids by colonic bacteria.

Our finding of higher fecal D- and L-lactate concentrations in diarrheic calves supports our hypothesis that bacterial fermentation of undigested and malabsorbed nutrients take place in the gut of diarrheic calves. The mechanisms by which these acids could be absorbed in diarrheic patients are poorly understood. There may be direct diffusion across inflamed mucosa. In addition, the mucosal monocarboxylic acid transporter transports L-lactic acid and a variety of other organic acids and transport is enhanced by low pH (40). Unlike L-lactate, metabolism of D-lactate in calves, humans and other vertebrates is slow (25) or impaired (41). Therefore, accumulation of D-lactic acid is an important contributory factor to acidosis in neonatal calves with diarrhea. Our study establishes that acidosis in diarrheic calves is due in part to D-lactic acid accumulation and that the gastrointestinal tract seems to be the site of D-lactate production. This finding was made possible by the new, fast and simple analytical approach used in this study, which made it possible to partition lactic acidosis between D and L isomers of lactic acid.