Changes of the erythrocyte phenotype and blood biochemistry in dairy calves during the first ten weeks of age

Introduction

Laboratory diagnostic tests can present a valuable tool for the production animal veterinarian to monitor herd health or to accurately diagnose early stage disease in individual animals. The meaningful and efficient use of laboratory results in calves requires a good understanding of physiological changes occurring as the animal grows. To improve the diagnostic precision of cell counts, cell indices or blood biochemical parameters, the establishment of reference intervals specific for age, gender, breed or production type has been encouraged in the literature (Dillane et al., 2018; Mohri, Sharifi & Eidi, 2007; Moretti et al., 2017; Panousis et al., 2018; Shaffer, Roussel & Koonce, 1981). Hematologic parameters of bovines such as the erythrocyte count, cell volume or the biochemical composition of erythrocytes are subject to important changes during the first months of life. Changes occurring in juvenile bovine erythrocytes include the replacement of fetal by adult hemoglobin (Hb), a reduction of the mean corpuscular volume (MCV), a reduction of the intracellular concentration of potassium ([K]) and the concomitant increase of the intracellular sodium concentration ([Na]) (Grimes, Duncan & Lassiter, 1958; Israel et al., 1972; Mohri, Sharifi & Eidi, 2007). Investigating these changes that occur in the first weeks of age is of interest as some parameters are used for example to diagnose metabolic disturbances of the iron homeostasis in calves, or potassium depletion in sick neonates (Bostedt et al., 2000; Bostedt, Jekel & Schramel, 1990; Golbeck et al., 2018; Volker & Rotermund, 2000; Wittek, Köchler & Mader, 2014). A decline of the MCV that physiologically occurs in early life is also considered a hallmark of iron deficiency anemia. For the accurate interpretation of laboratory results it is thus essential to differentiate between physiological and pathological changes (Constable et al., 2017). Exploring associations between some of these parameters may furthermore provide valuable clues for a better understanding of mechanisms through which such changes take place. The objectives of this study thus were to describe hematological and specific blood biochemical changes occurring in healthy Holstein-Friesian calves during the first 10 weeks of age, and to identify associations between these parameters possibly explaining underlying mechanisms.

Materials and Methods

Results

Average birth weight of experimental animals was 40.7  ± 4.5 kg (mean  ± SD). Only minor disease incidents were recorded during the study. Therapeutic intervention was required in the first and second week for two calves with mild omphalitis, for two calves in the seventh week, one calf in the ninth week for respiratory tract disease, and two calves in the fifth and seventh week that developed mild diarrhea. None of the calves required treatment for more than one condition or on more than one occasion during the study. Treatment for omphalitis and respiratory tract disease consisted of parenteral antimicrobial therapy (Amoxicillin, 10 mg/kg BW subcutaneously) in combination with a non-steroidal anti-inflammatory drug (Meloxicam, 0.5 mg/kg BW subcutaneously). Mild diarrhea was exclusively treated with a non-steroidal anti-inflammatory drug (Metamizole, 20 mg/kg BW intramuscularly). Treated animals made uneventful recoveries in all instances. All samples obtained at the time of disease remained in the dataset for statistical analysis as none of the parameters deviated more than 5% of the mean value of the sampling week without the sample in question. Due to clotting, 10 EDTA blood samples could not be analyzed. Intake of MR/acidified whole milk was consistently equivalent to the total amount of liquid nutrient offered throughout the study. Transient feed intake depression was recorded on the day calves were moved from hutches to group pens, which was primarily attributed to the change of feeding system and stress from relocation/grouping (Fig. S1).

Gender specific differences were not apparent for any of the studied parameters.

The time curves for RBC and MCV are presented in Fig. 1. While the RBC increased from 7.2 ± 1.1 × 10¹²/L to 9.3 ± 1 × 10¹²/L in the first five weeks of age and to 10 ± 0.7 × 10¹²/L until the end of the study, the MCV declined from 43.6 ± 3.7 fL to 35.6 ± 3.2 fL between the first and seventh week, but remained stable thereafter. This inverse pattern for RBC and MCV resulted in a constant PCV throughout the study (Table 1). The blood hemoglobin concentration increased from 0.96 ± 0.16 g/L to 1.16 ± 0.11 g/L during the first three weeks of age, without any further change during the remainder of the study period (Fig. 2). Highest plasma [Fe] were measured in the first week of age that then declined continuously, reaching a nadir at seven weeks of age. Thereafter an increase in mean plasma [Fe] was observed until the end of the study (Fig. 2; Fig. S3). Values for plasma [Fe] were highly variable within and between animals. Plotting the plasma [Fe] values measured in the first week of age against the age (in days) at sampling, revealed a decrease in plasma [Fe] during the first week of age (Fig. S3).

K_ERY and Na_ERY are depicted in Fig. 3. K_ERY declined from a peak value of 91.9 ± 13.5 mmol/L in the first week to a nadir of 24.6 ± 7.2 mmol/L at the end of the study. A concomitant increase of Na_ERY was observed from the mean nadir of 45.0 ± 32.0 mmol/L in the first week of age to a mean peak value of 106.1 ± 38.4 mmol/L measured at the eighth week of age. Although the change of Na_ERY over time mirrored the content-time curve of K_ERY, a larger degree of variation was observed for Na_ERY. Compared to the intracellular electrolyte contents, plasma [K] showed only a mild decline over time, and changes in plasma [Na] were insignificant (Table 1). This observation also reflects in calculated ECC values that remained constant over time. In contrast the ICC and thereupon the DIFF showed a larger degree of variation and subtle changes between weeks (Fig. S2).

The stepwise regression analysis with MCV as dependent variable including all sampling time points showed strongest associations with K_ERY (partial R² = 0.480; P < 0.0001). The correlation analysis stratified by sampling week revealed significant negative associations of the MCV with K_ERY during the last five weeks of the study, and a positive association with Hb during the last two weeks.

The stepwise regression analysis with Hb as dependent variable including all time points showed strongest associations with RBC (partial R² = 0.280; P < 0.0001). The regression coefficient however decreased over time in the analysis stratified by week of age (Table 2). The correlation analysis identified a negative association between Hb and K_ERY during the first and last two weeks of the experimental period, and a positive association between RBC and Hb over the entire study period (Table 3).

Apart from a positive correlation with K_ERY in the first and fourth week (Table 3) no other significant association with plasma [Fe] could be identified.

The stepwise regression analysis with K_ERY as dependent variable including all sampling times did not reveal associations with any of the included parameters (Table 2). The regression analysis with PCV as dependent variable only identified a significant association with Hb (partial R² = 0.368; P < 0.0001) (Table 2).

Discussion

The longitudinal observation of the hemogram and blood biochemical parameters over the first 10 weeks of age in dairy calves showed a considerable decline of erythrocyte volume occurring concomitantly with a marked decline of the erythrocyte K content, and a more moderate increase of the erythrocyte Na content. These results are in line with previously published values for MCV in calves of different age, and seem to be independent of breed, gender, or whether calves are fed milk replacer or whole milk (Benesi et al., 2012; Brun-Hansen, Kampen & Lund, 2006; Egli & Blum, 1998; Holman, 1956; Mohri et al., 2004; Mohri, Sharifi & Eidi, 2007; Moosavian, Mohri & Seifi, 2010; Tennant et al., 1973).

The observed decline in MCV is consistent with the development of microcytosis, which, if associated with hypochromia, is considered a hallmark of iron deficiency anemia. Because the reduction in red blood cell volume was accompanied by a more pronounced increase of the RBC, the plasma Hb increased over time. There was thus no indication of developing anemia in this study. These results show that a declining MCV in neonatal calves is a physiological process and should thus not per se be interpreted as sign of masked iron deficiency anemia. This conclusion is supported by the plasma [Fe] determined in this study that were above concentrations previously reported in neonatal calves and indicate that there was no sideropenia in study animals in the time range during which the decline of MCV occurred (Bostedt, Jekel & Schramel, 1990; Mohri et al., 2004; Mohri, Sharifi & Eidi, 2007).

The comparably high values of plasma [Fe] at the first sampling time may be attributed to the oral administration of an iron supplement on the first day of life. A similar observation was made in an earlier study (Kume & Tanabe, 1996). It should however be noted that the single dose of 230 mg Fe³⁺ administered orally is low compared to the recommend dose of 1,000 mg of Fe³⁺ per calf to be administered parenterally (Bostedt et al., 2000; Steinhardt & Thielscher, 2003). The declining plasma [Fe] during the first weeks of life reaching a nadir at seven weeks of age, and the following increase over the last weeks of the study suggest that MR as currently fed under field conditions, containing 100 mg Fe/kg DM, as in this study provides less Fe than what healthy calves consume after weaning.

The plasma [Fe] showed a remarkable intra- and interindividual variation, and although the decline in plasma [Fe] and MCV occurred in the same time interval the plasma [Fe] did not associate to any of the parameters of the hemogram. A correlation of plasma [Fe] with Hb has been reported with considerably lower plasma [Fe] but seems to vanish with higher plasma [Fe] (Wittek, Köchler & Mader, 2014). From our results we conclude that the important decline in plasma [Fe], reaching a mean nadir of just above 20 µmol/L at seven weeks of age did not result in measurable negative effects on hematological parameters. Because Fe is also required for myoglobin production and oxygenation of muscle fibers, effects on muscle growth and function must also be considered when assessing adequacy of the dietary Fe supply.

Normal values for hematology parameters of newborn calves published in earlier studies are inconsistent which seems attributable to differences between studies in gender, breed, sample size, feeding management and analytical techniques employed. The increase in Hb and RBC observed in the present study is consistent with values determined in iron supplemented calves (Mohri et al., 2004; Moosavian, Mohri & Seifi, 2010) but are also in the range of values determined in unsupplemented calves (Brun-Hansen, Kampen & Lund, 2006; Mohri, Sharifi & Eidi, 2007; Moosavian, Mohri & Seifi, 2010). The steady PCV during the first weeks of age described here is in agreement with some earlier studies (Benesi et al., 2012; Brun-Hansen, Kampen & Lund, 2006; Mohri et al., 2004; Panousis et al., 2018), while others report a considerable decline of the PCV over the first weeks of age (Egli & Blum, 1998; Holman, 1956; Tennant et al., 1973). Similarly the slight decline of the MCH matches some of the existing reports (Benesi et al., 2012) while other authors report stable values (Panousis et al., 2018; Tennant et al., 1973) or more pronounced changes (Mohri, Sharifi & Eidi, 2007). Constant MCHC values during the first weeks of age, as in this study have been reported (Holman, 1956; Panousis et al., 2018), as have increasing (Mohri et al., 2004) and decreasing MCHC values (Mohri, Sharifi & Eidi, 2007).

The intracellular osmotic and oncotic pressure, and the osmotic equilibrium between intra- and extracellular space are important determinants of intracellular volume. Osmotic forces are primarily driven by electrolytes, carbohydrate- and other metabolites that are present on both sides of the cell membrane (Lang et al., 1998). The electrolytes most relevant for the extracellular osmotic pressure are sodium followed by chloride, while the intracellular osmotic pressure strongly depends on the intracellular [K] and to a lesser degree the [Na]. Since erythrocytes lack cell organelles requiring or producing metabolites with osmotic activity, hemoglobin is the quantitatively most abundant protein inside the red blood cell that strongly influences the intracellular oncotic pressure and thereby the cell volume (Bogner et al., 1998; Bogner et al., 2002; Lang et al., 1998). These effects reflect in the associations of the MCV with K_ERY, DIFF and hemoglobin reported in this study.

K_ERY and Na_ERY underwent important changes over time, with K_ERY declining by over 60% during the study while Na_ERY more than doubled over the same period. Erythrocytes with different physiological [K] have been recognized in various species, and have been referred to as high potassium (HK) and low potassium (LK) phenotypes, depending on the physiological range in the specific individuum (Israel et al., 1972; Komatsu et al., 1980). Our results corroborate the findings of an earlier study based on single measurements in Holstein-Friesian calves of varying age and covering an age range of six months (Israel et al., 1972). Over the same time period only numerical decreases of the plasma [K] and a concomitant numerical and equimolar increase of the plasma [Na] were observed, with both electrolyte concentrations in plasma remaining within the reference ranges for calves of the studied age group (Dillane et al., 2018). As a result of the inverse changes in [Na] and [K] of the intra- and extracellular space, the calculated values ECC, ICC and DIFF remained constant, indicating that the observed changes occurred without measurably disturbing the osmotic equilibrium on the erythrocyte membrane. Interestingly, correlation- and regression analyses revealed consistent negative associations between MCV and K_ERY from the sixth week of age on, so at a time when both values approached levels of adult animals but not earlier. This finding indicates that the decline in red blood cell volume, although occurring simultaneously with changes in intracellular electrolyte concentrations is not causatively related to these.

The described changes in intracellular electrolyte content have been attributed to a reduction in the activity of the Na⁺-K⁺-ATPase (Bogner et al., 2002; Israel et al., 1972; Tosteson, 1966). A reduced activity of the Na⁺-K⁺-ATPase with a pumping ratio of potassium to sodium of 3:2 also provides a plausible explanation for the higher degree of variation observed for Na_ERY compared to K_ERY in this study (Post & Jolly, 1957).

It remains to be understood why considerably higher concentrations of erythrocyte K in neonatal erythrocytes do not seem to affect cell volume, while lower concentrations in adult erythrocytes do. Also, the purpose of this switch from HK to LK in the first weeks of age remains to be elucidated. As cell volume has a regulatory effect on a variety of cell functions, a possible explanation could be the concomitantly occurring switch from fetal to adult Hb in calves in the studied age range (Blechner, 1960; Lang et al., 1998). Studies conducted on human erythrocyte populations that were affected by mutations in the hemoglobin chain revealed that different hemoglobin morphologies influence the ion transport activity through the cell membrane by altering the K–Cl—cotransport, which affects the erythrocyte K content (Olivieri et al., 1992). After birth the fetal hemoglobin fraction diminishes and is entirely replaced by adult hemoglobin by eight to 13 weeks of age. This process is accompanied by the development of a new, adult, erythrocyte population, replacing the fetal population (Frei, Perk & Danon, 1963; Grimes, Duncan & Lassiter, 1958). In the present study the decline of fetal- and appearance of adult hemoglobin was not studied, which precluded the possibility to investigate the association between the change in type of hemoglobin and the change in MCV occurring in early life.

The available literature suggests that not all adult cattle display the LK erythrocyte phenotype. For bovines the erythrocyte phenotype has been suggested to vary between breeds (Table 4). LK and HK erythrocyte phenotypes have also been described in goats (Blechner, 1960; Evans & Phillipson, 1957), sheep (Blechner, 1961; Evans, 1957) and non-ruminant species (Beilin et al., 1966; Kerr, 1937). Studies conducted in Holstein-Friesian cattle as in the present study are scarce, but from available data it becomes apparent that the LK phenotype is at least highly predominant, and even the absence of HK phenotype in this breed has been suggested (Israel et al., 1972; Komatsu et al., 1980). Therefore, the consistent shift from a HK to the LK erythrocyte phenotype observed in our study population exclusively consisting of Holstein-Friesian calves appears plausible.

The variability of the erythrocyte phenotypes reported in the literature as well the changes occurring in K_ERY in early life present important caveats when considering the use of erythrocytes as easily accessible and processable cells to assess the total body- or intracellular K homeostasis, as this has been proposed in other species (Muylle et al., 1984). The results presented here suggest that erythrocytes would be particularly unsuitable as target cells to assess the K balance in the first 10 weeks of age, which can be of interest for example when treating calves with neonatal diarrhea (Golbeck et al., 2018).

As the objective of the study presented here was to describe physiological changes as the calf grows, particular care was taken to limit the bias by disease or unphysiological conditions on the investigated parameters to a minimum by closely monitoring study animals for early signs of clinical disease. Oral iron supplementation in moderate amounts at birth was included to limit hematological disturbances attributable to possible latent iron deficiency. The initially relatively high plasma [Fe] indicate that this objective was met. We therefore consider observed changes to be physiological developments corresponding to the age period of Holstein-Friesian calves.

Conclusions

Several important changes in erythrocyte phenotype and biochemical composition occur in bovine erythrocytes in the first 10 weeks of age. Observed changes in cell volume could neither be associated to the decline in erythrocyte K content nor the plasma [Fe], leaving the underlying mechanism behind cell volume reduction unidentified. The results presented here emphasize that the sole reduction of red blood cell volume in early life cannot be interpreted as symptom for iron deficiency anemia without taking into consideration the concomitantly occurring physiological changes. The important changes of erythrocyte K content occurring in bovine neonates make red blood cells particularly unsuitable for the reliably representation of the K content of the intracellular space.

Supplemental Information