Thirteen-lined ground squirrels (Spermophilus tridecemlineatus) exploit the low-temperature activity of pancreatic triacylglycerol lipase (PTL) during hibernation. Lipolytic activity at body temperatures associated with hibernation was examined using recombinant ground squirrel and human PTLs expressed in yeast. Both the human and ground squirrel enzymes displayed high activity at temperatures as low as 0°C and showed Q10 values of 1.2–1.5 over a range of 37–7°C. These studies indicate that low-temperature lipolysis is a general property of PTL and does not require protein modifications unique to mammalian cells and/or the hibernating state. Western blots show elevated levels of PTL protein during hibernation in both heart and white adipose tissue (WAT). Significant increases in PTL gene expression are seen in heart, WAT, and testes; but not in pancreas, where PTL mRNA levels are highest. Upregulation of PTL in testes is also accompanied by expression of the PTL-specific cofactor, colipase. The multi-tissue expression of PTL during hibernation supports its role as a key enzyme that shows high activity at low temperatures.
- low temperature
in certain climates, winter presents conditions unfavorable for survival. A decrease in ambient temperature and an inadequate food supply introduce a high energetic cost to mammals trying to maintain normal physiological functions (reviewed in Ref. 31). Many mammals prepare for these conditions by caching food or by increasing their food intake in the late summer and fall months, thus adding to their body fat stores. However, for some mammals, particularly those with a small body mass (<5 kg), these two behaviors are insufficient to maintain a constant body temperature for the whole of winter in the face of severe cold and no food. A mouse or similar-sized mammal, for example, would have to add 320% extra body mass just to maintain basal metabolic rate at a 37°C body temperature without food for a period of 100 days (31). On the other hand, an estimated 87.8% energy savings is gained in another rodent, Spermophilus richardsonii (Richardson’s ground squirrel), partaking in hibernation for one season (44). Therefore, one alternative for coping with the significant energetic cost of winter is to hibernate.
During hibernation, body temperature typically lies within a few degrees of ambient temperature and metabolic rate is maintained at ∼1–3% of the resting state under normal physiological conditions (reviewed in Ref. 27). In hibernating mammals, the respiratory quotient (RQ), or the ratio of the volume of carbon dioxide produced to the volume of oxygen consumed, hovers around 0.7, revealing that fats rather than carbohydrates are the major source of fuel in this reduced metabolic state (reviewed in Refs. 27 and 39). A focus of our research has been the identification of genes responsible for metabolic changes that occur during hibernation in mammals (1, 7).
We previously reported that a particular enzyme, pancreatic triacylglycerol lipase (PTL), was expressed in the heart and white adipose tissue (WAT) of the thirteen-lined ground squirrel (S. tridecemlineatus) during hibernation (1, 3). Until recently, PTL was thought to be expressed exclusively in the pancreas and secreted into the small intestine where it digests dietary fat (reviewed in Ref. 24). However, we found that soluble protein extracts made from hibernating ground squirrel heart and WAT exhibited measurable PTL activity at 37°C. Furthermore, when activity measurements were taken over a broad temperature range, PTL expressed in the heart exhibited significant lipolysis at near-freezing temperatures (1). Taken together, these findings suggested that PTL provides one mechanism for liberating the fatty acids required in this species during hibernation.
In this paper we further examine the role of PTL as an important lipolytic enzyme during hibernation. First, we express both recombinant human and thirteen-lined ground squirrel PTL enzymes in the yeast, Pichia pastoris. We then measure their activities against a purified triacylglycerol and lipids extracted from ground squirrel WAT across a temperature range from 0 to 37°C. Next, we explore the relationship between PTL protein levels and enzymatic activity in heart and WAT to determine whether differences in activity levels are a function of the amount of protein present in these two tissues at various times throughout the year. We also characterize the expression of the PTL-specific cofactor, colipase, and suggest that PTL functions in the heart and WAT of the hibernating ground squirrel without the aid of this protein. Last, we show that PTL and colipase are expressed in parallel during hibernation in another novel location, the testes.
MATERIALS AND METHODS
The animal care and use described below was approved by the Institutional Animal Care and Use Committee. Both male and female thirteen-lined ground squirrels (S. tridecemlineatus) were obtained each August from TLS Research (Bartlett, IL) within 3–4 days of wild capture or were live-trapped in Mount Pleasant, MI. Squirrels were kept individually in plastic top-load rat cages filled with pine shavings. Diet consisted of standard laboratory rodent chow (Purina, no. 5001) and was supplemented with black oil sunflower seeds and water ad libitum. Food availability, room temperature, and photoperiod were varied artificially to induce hibernation in captivity. From August to mid-November squirrels were maintained on a 12:12-h light/dark cycle with food and water available ad libitum. During this time room temperature was stepped down gradually. In August it was maintained at 23°C, in September at 17°C, in October at 11°C, and from mid-November to mid-March at 5°C. In mid-November, food was removed, and squirrels were housed in complete darkness with only water ad libitum. Under these conditions, all animals hibernated. To induce terminal arousal, in mid-March, the room temperature was increased to 23°C, food was once again made available, and a 12:12-h light/dark cycle was reinstated.
From October to mid-March, the state of the squirrel (active vs. hibernating) was monitored using the sawdust technique (32). Organ and tissue collection times were chosen to represent different states of animal activity combined with various ambient temperatures, months, and/or food availability. A naturally occurring activity state referred to as an interbout arousal (IBA) indicates an animal that, on the day of death, was active during the hibernation season and had experienced at least three previous hibernation bouts of 8 days or more. Body temperatures of hibernating squirrels were measured rectally before death. Squirrels were killed by decapitation, and body temperatures of active animals were measured rectally immediately after death. Upon removal from the animal, each organ or tissue was placed in a cryogenic tube and then immediately submerged in liquid nitrogen. If RNA was to be made from the pancreas, then the tissue was placed directly in 4 M guanidinium isothiocyanate and homogenized immediately upon removal from the animal, because of the observed lability of RNA prepared from ground squirrel pancreas. For long-term storage, organs and tissues that were quick-frozen were placed in liquid nitrogen tanks or in a −80°C freezer.
Northern blot analysis.
RNA was isolated as described in the accompanying paper (40). Total RNAs were separated electrophoretically, transferred to a nylon membrane, hybridized to 32P-labeled probes, and quantified as described previously (1). Pancreatic and testes RNA blots used to examine PTL and colipase expression contained 10 μg total RNA per lane, as did the multi-tissue blot used to examine colipase expression in 10 hibernating ground squirrel tissues. The PTL probe employed in these studies was generated from the EcoRI-XbaI fragment of a cDNA clone isolated from a ground squirrel hibernating heart cDNA library (accession no. AF027293). This fragment contains the majority of the PTL coding region. The full-length colipase cDNA (accession no. AF395869) utilized to generate the probe for hybridization was isolated from a ground squirrel pancreas cDNA library (described below). To control for RNA integrity and loading, all blots were hybridized separately with a probe complementary to the full-length human β-actin cDNA (accession no. X00351; Ref. 33) and with an oligonucleotide probe complementary to 18S rRNA (5′ CGACTTTTACTTCCTCTAGATAGTCAAGTTCGAC 3′). The cDNAs and the PTL cDNA fragment were labeled via random priming, as first described by Feinberg and Vogelstein (12, 13), using the Rediprime II random primed labeling system (Amersham). This reaction was allowed to proceed at 37°C for 30 min. The 18S rRNA oligonucleotide was end-labeled as described in Sambrook et al. (36).
PTL and colipase mRNA values for each lane were normalized to their respective 18S rRNA values. After normalization, relative values were determined by comparison to the average summer (ACTIVE-Aug) value. This was done separately for PTL and colipase bands. One-way ANOVAs were performed, and P values were computed using Microsoft Excel 2000.
Pancreas cDNA library construction and screening.
A thirteen-lined ground squirrel pancreas cDNA library was constructed using mRNA from a hibernating and an active thirteen-lined ground squirrel pancreas so that both activity states could be represented in a single library (7). The library was screened initially for PTL using a 32P-labeled EcoRI fragment of a cDNA clone isolated from a ground squirrel hibernating heart cDNA library (accession no. AF027293). This fragment contained the first 600 bases of the heart PTL mRNA. To isolate full-length clones and to remove bias from the library screen that would direct the discovery of PTL cDNAs that resembled those found in the heart, the library was rescreened for PTL using a 32P-end-labeled oligonucleotide (5′ AGCAGCAGTGCCAGCGACCAGACCAGCAGC̅A̅T̅CATG 3′) containing the complement of the start codon (underlined) at its 3′ end. To isolate full-length colipase cDNA clones, the library was screened with a 32P-end-labeled colipase oligonucleotide (5′ GAGCACTCGCTGTTCTCGCTGGCCTTGGVTGTGCAGCGGGCCAGGC 3′, where V is a degenerate site and could be an A, C, or G). This oligonucleotide was designed from a multi-sequence alignment of mammalian colipase orthologs. cDNA clones were sequenced using ABI Prism 377 automated cycle sequencers (PE Applied Biosystems). Finished sequences were compared with known sequences entered into the National Center for Biotechnology Information (NCBI) database using BLAST (4).
This procedure is described in the accompanying paper (40). To measure colipase expression, the PCR program was allowed to proceed for 28 cycles, and to measure actin expression, it proceeded for 20 cycles. Denaturation was performed at 94°C for 30 s, annealing at 55°C for 45 s, and extension at 72°C for 1 min. The primer pair used to measure colipase expression was 5′ TGGCCTATGCAGCTCCTG 3′ and 5′ ACAGGTCAGGCCMCGCTC 3′, where M is a degenerate site that could be an A or C. The primer pair used to measure β-actin expression was 5′ GACAGGATGCAGAAGGAG 3′ and 5′ ACATCTGCTGGAAGGTGG 3′. PCR results were viewed on 5% acrylamide, 1× TBE gels.
Several controls, as described in the companion paper (40), were enlisted to ensure that the products generated from the PCR reactions represented message expressed in a particular organ or tissue and were not the result of genomic DNA or other DNA contaminants. The primer pair selected to amplify the colipase message was chosen so that forward and reverse primers spanned at least one intron based on the nucleotide sequence for the human colipase gene (accession no. M95529).
Expression and isolation of recombinant PTL protein.
Recombinant ground squirrel PTL and recombinant human PTL were expressed in P. pastoris and isolated from this system as described in Yang and Lowe (45). To accomplish this task, human PTL cDNA (accession no. NM_000936) and ground squirrel heart PTL cDNA (accession no. AF027293) were subcloned into pHIL-S1 expression vectors and transformed into GS115 yeast cells. This subcloning process replaced the PTL signal peptide with a functional yeast PHO1 signal peptide. As a result, the PTL protein that was secreted from the expressing yeast cells into the surrounding media represented the mature 449-amino acid protein. As checkpoints for this procedure, both cDNA inserts were sequenced after subcloning into the pHIL-S1 expression vectors. In addition, the identity of the recombinant human PTL protein was verified by amino-terminus sequencing of the purified protein (45).
Extraction of lipids from WAT.
WAT from three hibernating thirteen-lined ground squirrels, one of which was killed in December (body temperature or Tb = 5.7°C) and the other two of which were killed in March (Tb = 6.4°C and 6.7°C, respectively), was pooled. Lipids were extracted from this pooled sample using a modification of the method outlined in Schwarz et al. (38). The starting volume of chloroform:methanol (2:1) used for extraction was 40 ml and was scaled up accordingly. The combined lipids extracted from the WAT of these hibernating animals were used as one of the substrates against which recombinant PTL activity was measured.
Recombinant human and ground squirrel PTLs were assayed for PTL activity using the pH-STAT method as outlined in Lowe (22). Assays were performed for 5 min. Initially, each recombinant enzyme was assayed in the presence of various concentrations of taurodeoxycholate, ranging from 0 to 4 mM, and with or without a fivefold molar excess of recombinant human colipase. These assays were performed at 37°C using 4-carbon chain tributyrin (Fluka) as the substrate. Later, assays were performed in the presence of 4 mM taurodeoxycholate, using 1.4 μg of recombinant PTL and a fivefold molar excess of recombinant human colipase. Using these conditions, PTL activity was measured against 18-carbon chain triolein (Sigma) and lipids extracted from the WAT of hibernating thirteen-lined ground squirrels. Assays were performed at five temperatures: 0, 7, 17, 27, and 37°C. The colipase was isolated as described in Cordle and Lowe (9). Emulsified substrates were prepared as described in Lowe (22). The data are presented as averages of two independent trials.
Soluble protein extracts were prepared from thirteen-lined ground squirrel heart acetone powders and were assayed for PTL activity as described in Andrews et al. (1). Soluble protein extracts were prepared from thirteen-lined ground squirrel WAT acetone powders and were assayed for PTL activity as described in Bauer et al. (3). Western blot analysis was performed as described in Buck et al. (7) with the exception of 150 mM NaCl in the TBST solution. PTL was detected using a 1:5,000 dilution of rabbit polyclonal antisera raised against recombinant thirteen-lined ground squirrel PTL. For both heart and WAT data, one-way ANOVAs were performed, and P values were calculated using Microsoft Excel 2000.
Recombinant PTL activity.
As previously reported in Andrews et al. (1), PTL from hibernating heart extracts exhibited significant lipolytic activity at temperatures as low as 0°C. We wanted to determine whether this cold-adapted activity was unique to PTL expressed in the hibernator. To test this hypothesis, cDNAs encoding ground squirrel heart PTL (accession no. AF027293) and human pancreatic PTL (accession no. NM_000936) were expressed in the yeast, P. pastoris (45), and lipolytic activity of these recombinant proteins was assayed over a temperature range from 37 to 0°C. Expression of both PTLs in the same yeast background averts the likelihood that a species-specific, tissue-specific, or hibernation-specific posttranslational modification affects enzyme activity. PTL cDNAs from heart, WAT (accession nos. AF177402 and AF177403), and pancreas (accession no. AF395870) encode identical amino acid sequences, and therefore the primary structure of the recombinant heart PTL expressed in yeast is the same for all three tissues. The size of the recombinant protein is also the same as that from various thirteen-lined ground squirrel tissues (40).
Recombinant human and ground squirrel PTLs were assayed in three separate experiments using the pH-STAT method (22) to measure their activities. To confirm that the ground squirrel protein is indeed PTL, the first experiment used the 4-carbon chain triacylglycerol substrate, tributyrin, in the presence of the inhibitory bile salt taurodeoxycholate and the PTL-specific coactivator colipase. The second experiment used the purified 18-carbon chain triacylglycerol, triolein, to test the effectiveness of both the human and ground squirrel enzymes on a purified substrate commonly found in hibernators (8, 15–17). Finally, to show the effectiveness of PTL against a heterogeneous mixture of triacylglycerols stored in the hibernator, lipids extracted from the WAT of hibernating thirteen-lined ground squirrels were used as the substrate in the third experiment. The last two experiments were conducted over a broad temperature range.
As seen in Table 1, recombinant thirteen-lined ground squirrel heart PTL expressed in yeast exhibited properties commonly reported for tissue-purified PTL or PTL purified from pancreatic juice (6, 19, 30). At an assay temperature of 37°C, a 0.5 mM concentration of the bile salt taurodeoxycholate was found to mildly stimulate PTL activity in the presence or absence of colipase. However, above the critical micelle concentration for taurodeoxycholate (0.9–1.3 mM), PTL activity was observed only in the presence of colipase. Thus recombinant ground squirrel PTL expressed and purified from P. pastoris shows inhibition of lipolytic activity by bile salt micelles, and this inhibition is overcome by the presence of colipase. Nearly identical results were obtained with the purified recombinant human PTL using these same assay conditions.
The specific activity profiles of recombinant ground squirrel and human PTLs using triolein as a substrate were remarkably similar over a range of temperatures (Fig. 1A). In the presence of colipase and taurodeoxycholate, ground squirrel PTL showed 38% maximal activity at a typical hibernating temperature of 7°C and 33% maximal activity at 0°C. Lipolytic activity in the absence of colipase and taurodeoxycholate was reduced, but remained robust at each temperature, showing 32–33% maximal activity at 7°C and 23–24% maximal activity at 0°C for both human and ground squirrel PTLs. Lipolytic activity against triacylglycerols extracted directly from WAT of hibernating squirrels was almost identical for ground squirrel and human PTLs across the entire temperature range in the presence of colipase and taurodeoxycholate (Fig. 1B). In this case percent maximal activities for PTL from both species were 53–54% at 7°C and 40% at 0°C. Similar values were seen in the absence of colipase and bile salts at 7°C, with 56% maximal activity for ground squirrel and 53% for human PTL. At 0°C these values were reduced to 46% for ground squirrel and 40% for human.
The cold-adapted character of PTL was also confirmed by calculating the Q10 (fold reduction in two reaction rates 10°C apart) over a range of body temperatures typically seen in mammals entering hibernation. In the presence of colipase and taurodeoxycholate, both ground squirrel and human PTLs exhibited low Q10 values of 1.3 over a range of 37–7°C against triolein. Over the same temperature range and reaction conditions, Q10 values for both human and ground squirrel PTLs were 1.2 against the WAT triacylglycerols. In the absence of colipase and taurodeoxycholate, the Q10 values using triolein as substrate were 1.5 for ground squirrel PTL and 1.4 for human PTL over a range of 37–7°C. Under these same conditions both ground squirrel and human PTLs exhibited a Q10 value of 1.2 using triacylglycerols extracted from WAT of hibernating ground squirrels as substrate. Establishing the cold-adapted character of PTL in the absence of colipase and bile salts is important for understanding the role of PTL during hibernation, because key ground squirrel tissues such as heart and WAT do not contain colipase (this study) and presumably do not contain bile salts.
PTL protein expression in thirteen-lined ground squirrel heart.
PTL protein expression in the heart of the thirteen-lined ground squirrel was examined at various times throughout the hibernation season. Western blots were made using soluble protein extracts prepared from three summer-active (August), three hibernating (December-January), and three spring-active (May) heart acetone powders. In active August animals, PTL protein levels in the heart were low but then increased in December and January when the animals were hibernating (Fig. 2A). This pattern of PTL protein expression resembled that seen for PTL mRNA and enzymatic activity in the heart during the same time period (1). Figure 2A shows that PTL levels seen during hibernation decline after spring arousal in May. Weak bands seen on the protein blots were seen with the preimmune serum and did not represent other forms of the PTL protein (data not shown).
To explore in greater detail the relationship between PTL protein and activity levels in the thirteen-lined ground squirrel heart, quantitative Western blot analysis was carried out using the same five summer-active (August) and five hibernating (December-January) heart samples employed for PTL activity analysis at 7°C and 37°C in Andrews et al. (1). Results indicated that average PTL protein levels were nearly 12-fold greater in hibernating animals (P = 0.0002) than in their summer-active counterparts (Fig. 2B). A comparison of PTL protein levels with their respective activity levels at 7°C and 37°C for both groups indicated a positive correlation between protein concentration and activity. At both temperatures, this relationship was statistically significant in summer-active (P ≤ 0.01 at 37°C, and 0.01 < P < 0.05 at 7°C) and hibernating (0.01 < P < 0.05 at 37 and 7°C) ground squirrel heart extracts. Therefore, the data are consistent with a positive association between the level of PTL protein found in the heart and the level of PTL activity observed at 37°C and 7°C (Fig. 2C).
PTL protein expression in thirteen-lined ground squirrel WAT.
PTL protein expression in WAT of thirteen-lined ground squirrels was examined in a manner similar to that described for the heart. Western blot analysis of soluble protein extracts prepared from three summer-active (August), three hibernating (December-January), and three spring-active (March) WAT acetone powders revealed elevated levels of PTL protein in the summer-active and hibernating squirrels compared with the spring-active animals (Fig. 3A). Although this pattern of protein expression was different from that seen in the heart, it did match the pattern of mRNA levels and enzymatic activity observed previously in thirteen-lined ground squirrel WAT (3).
To gain a more detailed look at PTL protein expression in WAT throughout the hibernation season, and to determine whether PTL activity in WAT was indicative of the level of PTL protein found in this tissue, quantitative Western blot analysis was performed on the same 30 WAT extracts used to determine PTL activity at 37°C in Bauer et al. (3). As shown in Fig. 3B, changes in the relative levels of WAT PTL protein expression were on the same order of magnitude as changes in WAT PTL mRNA expression and enzymatic activity (3). Compared with the March-active values, PTL protein levels in August- through October-active animals were 2.5- to 3.8-fold higher. Only the level in the October-active group, however, was significantly higher (P = 0.04). During hibernation, including during IBAs, protein levels varied from 1.2- to 1.9-fold greater than spring levels, although these increases were not significant.
Colipase expression in hibernating thirteen-lined ground squirrel tissues.
To further assess the function of PTL during hibernation, it was necessary to characterize the expression of the PTL-specific cofactor, colipase. In its well-characterized role of digesting dietary lipids, PTL and colipase are expressed in pancreatic acinar cells (25, 26) and are secreted simultaneously into the duodenum of the small intestine (11). Here, they mix with the products of gastric lipolysis, bile secretions, and other enzymes that are present in the pancreatic juice. However, in hibernating thirteen-lined ground squirrels, we found various levels of PTL mRNA expression in every tissue examined (40). Therefore, it was important to determine whether colipase mRNA was also seen in these same tissues during hibernation.
To study colipase expression, full-length colipase cDNAs were also isolated from a thirteen-lined ground squirrel pancreas cDNA library. The 562-base consensus sequence derived from these colipase cDNAs (accession no. AF395869) encoded a 111-amino acid protein with a predicted molecular mass of 12,055 Da. This unprocessed form of ground squirrel pancreatic colipase, termed preprocolipase, shared 78% amino acid identity with human (accession no. NP_001823) and rat (accession no. NP_037271) pancreatic preprocolipases and 78% amino acid identity with mouse stomach preprocolipase (accession no. NP_079745). Northern blot analysis of 10 hibernating ground squirrel tissues probed with an isolated ground squirrel colipase cDNA revealed a single band present in the pancreas lane (Fig. 4A). The size of this message, estimated using RNA molecular weight markers, was 550 bases, which is in close agreement with the length of the cDNA clones isolated from the pancreas cDNA library. Further study, using RT-PCR on these same RNA samples, uncovered low-level expression of colipase in testes as well (Fig. 4B).
PTL and colipase expression in pancreas throughout the hibernation season.
PTL and colipase mRNA levels in the pancreas varied little from August through June (Fig. 5). Expression remained relatively constant even during the months from November through March when the animals were not eating. Additionally, when slight changes in the expression patterns of PTL and colipase were seen, these tended to show a similar pattern for both messages (Fig. 5B). At their highest, mean PTL mRNA levels for certain categories were only 1.5-fold greater than the mean summer-active PTL level (August), while mean colipase mRNA levels were ∼1.4- to 1.9-fold greater than their corresponding mean summer-active level (August). These nonsignificant elevations occurred in fall-active (Sep-Nov), fall-hibernating (Oct- Nov), and spring active (Mar-Jun) animals for PTL, and in fall-active (Sep-Nov) and spring active (Mar-Jun) animals for colipase. The one significant change in expression was observed with colipase in the IBA group. Relative to summer, colipase mRNA levels were down 30% during IBAs (P = 0.03). A Northern blot representing these patterns of expression across the hibernating season is presented in Fig. 5A. Reduced levels of expression seen in the August-active and March-hibernating points are due to lower amounts of RNA in those lanes as judged by the 18S rRNA profile shown below the blot.
PTL and colipase expression in testes throughout the hibernation season.
Unlike in the pancreas, PTL mRNA levels in testes were significantly increased during hibernation (P = 0.009) and during IBAs (P = 0.04) compared with the mean August value (Fig. 6C). These levels started their increase in the fall while the animals were active and remained high throughout the hibernation season, including during IBAs, before decreasing again in the spring after terminal arousal (Fig. 6A). This pattern resembled the pattern of PTL expression seen in the heart (1). In that case, however, the normalization was done with actin as a loading control. The 18S rRNA levels were used in this study because the actin levels in testes varied across the hibernation season (Fig. 6A).
With respect to colipase expression in the testes, it was not possible to determine relative levels of mRNA expression using Northern blot analysis. The colipase message was weakly detected after a several-day exposure, which enabled background levels to increase (Fig. 6A). As a result, RT-PCR was employed to determine whether colipase message was expressed in testes throughout the entire hibernating season. Using this method, colipase expression was seen in the testes of all animals except the one killed in August (Fig. 6B).
PTL mRNA is upregulated during hibernation in heart, WAT, and testes of the thirteen-lined ground squirrel, S. tridecemlineatus, as shown in this report and others (1, 3). Therefore, differential expression of PTL is seen in three nontraditional locations, but we found little change in PTL mRNA levels in the traditional organ of synthesis, the pancreas (Fig. 5). This pattern of PTL expression in hibernating squirrels supports the hypothesis that the hibernating phenotype arises from the differential expression of existing mammalian genes, rather than the evolution of genes that are specific to hibernation (41). In this paper we detail the cold-adapted properties of PTL (Fig. 1) and show that seasonal changes in thirteen-lined ground squirrel heart and WAT PTL mRNA levels lead to coordinate changes in protein levels (Figs. 2 and 3). Precise measurements of the absolute levels of PTL in heart and WAT were not possible due to technical reasons. However, we show that relative increases in PTL protein during hibernation correlates with a net increase in PTL enzymatic activity in heart and WAT (1, 3).
In vitro hydrolysis of triacylglycerols by PTL is inhibited by the presence of bile salts at certain concentrations (6, 19, 30). However, if colipase is added to these preparations, then PTL is stabilized and its lipolytic activity is restored (6, 30). In previous studies we added bile salts and colipase to heart and WAT extracts to distinguish PTL activity from that of other enzymes that hydrolyze triacylglycerols (1, 3). In the duodenum, where bile salts are naturally present, PTL is dependent on colipase to achieve its maximal activity. Therefore, it is not surprising that colipase mRNA is not found in heart and WAT where bile salts are not present (Fig. 4). We show in Fig. 1 that the absence of bile salts and colipase does not eliminate PTL activity or its ability to hydrolyze triacylglycerols at the low temperatures seen during hibernation.
The expression of PTL in testes (Fig. 6) could liberate fatty acids to be used as fuel for testes growth and recrudescence during hibernation. The pattern of involution, quiescence, and recrudescence of gonads in hibernators has been described by Blake (5). A study conducted by Barnes and colleagues (2) examined the timing of testes growth in male S. lateralis. They found the testes in this ground squirrel species doubled in size during the hibernation season. The cold-adapted activity of PTL could provide fuel for this growth during, and immediately after, hibernation in thirteen-lined ground squirrels. However, it is unclear why colipase is expressed in testes, because it is unlikely that bile salts or other emulsifying agents are present in gonadal tissue. Expression of colipase could result from a reported transcriptional promiscuity in testes due to high levels of the RNA polymerase II machinery in rodent spermatids (reviewed in Ref. 37).
In nonhibernating species, extra-pancreatic expression of PTL has been documented in the brain (42) and intestine (28) of rats. In rat intestine, dietary lipid raises PTL mRNA levels followed by an increase in PTL enzymatic activity (28). Recently, high-fat feeding in mice has been shown to increase PTL expression in the pancreas (35). In neonatal rat heart cells, long-chain fatty acids palmitic and oleic acid have been shown to induce the expression of genes coding for proteins involved in fatty acid transport and catabolism (43). Upregulation of PTL by dietary lipid may explain the timing of PTL expression in thirteen-lined ground squirrels. For example, heart and WAT PTL mRNA are highest during the fattening period just prior to hibernation (1, 3). Testes levels peak slightly later at the start of hibernation (Fig. 6). Increases in dietary fat consumption, combined with a putative promoter/enhancer activity derived from insertion of an endogenous retroviral sequence upstream of the PTL gene (40), are potential contributing factors to PTL expression in novel tissues during hibernation.
An obvious explanation for the expression of PTL in several tissues during hibernation is that it retains high activity at low temperatures. We previously showed that extracts containing PTL from the heart of hibernating thirteen-lined ground squirrels showed 61% and 34% maximal colipase-dependent activity at 7°C and 0°C, respectively (1). To determine whether this cold lipolysis is unique to ground squirrel PTL, or possibly the result of a hibernation-specific modification of the lipase, we expressed cDNAs for both human and thirteen-lined ground squirrel PTLs in the yeast P. pastoris (45). Expression of the two recombinant PTLs produced identical-length 449 amino acid proteins lacking the 16 amino acid N-terminal signal peptide. Because both lipases were expressed in the same yeast background, we avoided posttranslational modifications that are unique to a specific mammalian tissue and/or state of animal activity. The processed human and ground squirrel enzymes share 86% amino acid identity. The specific activities of these recombinant enzymes measured against both triolein and lipids extracted from hibernating ground squirrel WAT were comparable across a temperature range from 37 to 0°C (Fig. 1).
The physiological significance of PTL during hibernation is seen in its catalytic properties and pattern of expression. This significance is underscored when compared with the enzyme typically responsible for intracellular hydrolysis of triacylglycerols, hormone-sensitive lipase (HSL). Expression of the gene for PTL during hibernation in thirteen-lined ground squirrels was found to be considerably higher than HSL in heart, pancreas, and testes (3). More importantly, the catalytic properties of PTL offer various advantages over HSL during the physiological extremes of hibernation. Figure 1 shows that PTL has robust lipolytic activity over a wide range of temperatures, even in the absence of cofactors such as bile salts and colipase (Fig. 1). This result demonstrates the versatility of PTL under various molecular and thermodynamic conditions that may be encountered during hibernation. Low-temperature lipolysis has also been reported for HSL in rats and humans (21, 34); however, HSL activity varies depending on insulin and catecholamine levels (10, 20), whereas PTL activity is hormone independent. Lipolytic assays at 37°C show PTL activity against triacylglycerol substrates (45) is at least 75-fold higher than HSL activity (18).
HSL activity during hibernation has been reported in WAT of golden-mantled ground squirrels (14) and in jerboa (29). Both genes for PTL and HSL are expressed at high levels in WAT of thirteen-lined ground squirrels during hibernation (3, 40). Due to the distinct catalytic and regulatory properties of PTL and HSL, this dual triacylglycerol lipase system provides a means by which the fuel requirements of hibernating thirteen-lined ground squirrels can be met without interruption. Figure 1 shows that neither mammalian, tissue-specific, nor hibernation-specific modifications of PTL are necessary for high lipolytic activity at low temperatures. Resistance to the cold, a characteristic of the ground squirrel PTL initially described in hibernating hearts (1), is thus a property common to the mature form of PTL in both thirteen-lined ground squirrels and humans.
We are grateful to Jayita Guhaniyogi for construction of the ground squirrel pancreas cDNA library. This paper was improved by the comments of M. Tredrea and K. Russeth.
This work was supported by US Army Research Office Grant DAAD19-01-1-0014 and by Augmentation Award for Science and Engineering Training DAAG55-97-1-0175.
Article published online before print. See web site for date of publication (http://physiolgenomics.physiology.org).
Address for reprint requests and other correspondence: M. T. Andrews, Dept. of Biochemistry and Molecular Biology, Univ. of Minnesota School of Medicine, 1035 University Drive, Duluth, MN 55812 (E-mail:).
- Copyright © 2003 the American Physiological Society