Neutral amino acid transport in bovine articular chondrocytes

Abstract

1. The sodium-dependent amino acid transport systems responsible for proline, glycine and glutamine transport, together with the sodium-independent systems for leucine and tryptophan, have been investigated in isolated bovine chondrocytes by inhibition studies and ion replacement. Each system was characterized kinetically.

2. Transport via system A was identified using the system-specific analogue α-methylaminoisobutyric acid (MeAIB) as an inhibitor of proline, glycine and glutamine transport.

3. Uptake of proline, glycine and glutamine via system ASC was identified by inhibition with alanine or serine.

4. System Gly was identified by the inhibition of glycine transport with excess sarcosine (a substrate for system Gly) whilst systems A and ASC were inhibited. This system, having a very limited substrate specificity and tissue distribution, was also shown to be Na⁺ and Cl⁻ dependent. Evidence for expression of the system Gly component GLYT-1 was obtained using the reverse transcriptase-polymerase chain reaction (RT-PCR).

5. System N, also of narrow substrate specificity and tissue distribution, was shown to be present in chondrocytes. Na⁺-dependent glutamine uptake was inhibited by high concentrations of histidine (a substrate of system N) in the presence of excess MeAIB and serine.

6. System L was identified using the system specific analogue 2-aminobicyclo(2,2,1)heptane-2-carboxylic acid (BCH) and D-leucine as inhibitors of leucine and tryptophan transport.

7. The presence of system T was tested by using leucine, tryptophan and tyrosine inhibition. It was concluded that this system was absent in the chondrocyte.

8. Kinetic analysis showed the Na⁺-independent chondrocyte L system to have apparent affinities for leucine and tryptophan of 125 ± 27 and 36 ± 11 μM, respectively.

9. Transport of the essential amino acids leucine and tryptophan into bovine chondrocytes occurs only by the Na⁺-independent system L, but with a higher affinity than the conventional L system.

Chondrocytes are remarkable because they can produce and maintain the orderly form of the cartilage matrix, primarily composed of collagen and proteoglycan, while being relatively isolated from vascular and neuronal influences. Although few in number, chondrocytes are the only cells available to adapt cartilage to local change (Green, 1971; Kuettner et al. 1982; Hall et al. 1996). Despite their obvious importance for the synthesis and maintenance of the extracellular matrix, the processes involved in substrate uptake by chondrocytes have still not been fully elucidated. In particular, amino acids are needed to synthesize the major components of the extracellular matrix, but very little research has been undertaken to characterize amino acid transport in chondrocytes.

The Na⁺-dependent transport systems for neutral amino acids that have been identified in mammalian cells include systems A, ASC, B^{0, +}, N and Gly (Barker & Ellory, 1990a; Hediger et al. 1995; Moseley, 1996; Devés & Boyd, 1998). Both systems A and ASC have been found to have ubiquitous tissue distribution with the exception that erythrocytes and reticulocytes lack system A (Guidotti et al. 1978). System Gly has been identified in hepatocytes (Christensen & Handlogten, 1981), erythrocytes and reticulocytes (Ellory et al. 1981). System B^0,+ was first described in mouse blastocysts (Van Winkle et al. 1988) and has been characterized in many vertebrate epithelial cells. System N has been identified in hepatocytes (Kilberg et al. 1980), human erythrocytes (Ellory & Osotomehin, 1983), skeletal muscle (Hundal et al. 1986) and murine P388 leukaemic cells (Lazarus & Panasci, 1986). However, there is increasing evidence for the presence of several Na⁺-independent amino acid transport systems in mammalian cells which dominate uptake of certain amino acids. Until recently system L was taken to be the only Na⁺-independent transport system for neutral amino acids (Weissbach et al. 1982). Other Na⁺-independent systems now found in certain mammalian cells include systems L1, L2, asc₁, asc₂, y⁺, T, b^{0, +} and C (Barker & Ellory, 1990; Castagna et al. 1997; Devés & Boyd, 1998). Therefore it would be premature to nominate system L as the system responsible for transport of an amino acid simply because uptake follows Michaelis-Menten kinetics in the absence of Na⁺ ions.

In the present study, the Na⁺-dependent amino acid transport systems in the bovine chondrocyte that are responsible for proline, glycine and glutamine influx have been investigated using specific inhibitors and substrate analogues, ion dependence and kinetic characterization. These three amino acids were selected because their abundance in the proteins of cartilage makes their uptake by chondrocytes important for function. In addition, their characterization may allow the wider identification of amino acid transport systems that have been previously restricted both in tissue distribution and substrate specificity (e.g. system Gly and system N). The Na⁺-independent neutral amino acid transport systems present in the bovine chondrocyte have been identified using the two essential amino acids leucine and tryptophan. Leucine uptake by mammalian cells has been confined to system L, but separation into systems L1 and L2 has been reported in the rat hepatocyte (Weissbach et al. 1982). Tryptophan had originally been reported to be restricted to transport via system L but subsequent studies in the human erythrocyte have identified system T as a route for aromatic amino acid transport (Rosenberg et al. 1980; Rosenberg, 1981). This system has also been reported in isolated rat hepatocytes (Salter et al. 1986) but is absent from Ehrlich ascites tumour cells (López-Burilla et al. 1985). It has been suggested recently that thyroid hormones such as T₃ are its natural substrates (McLeese & Eales, 1996).

METHODS

Media

Cartilage slices and chondrocytes were maintained in Dulbecco's modified Eagle's medium (DMEM; Gibco), containing 5.5 mM glucose and buffered with 25 mM Hepes, to which was added 5 % (v/v) fetal calf serum (FCS), plus gentamicin and penicillin (both at 33.3 μg ml⁻¹). The salt composition was as follows (mM): NaCl, 109.5; KCl, 5.4; CaCl₂, 1.8; NaH₂PO₄, 1.04; MgSO₄, 0.81. The pH was adjusted to 7.4 with concentrated sodium hydroxide.

Preliminary washing and incubation of chondrocytes was performed in isosmotic saline medium buffered to pH 7.4 at room temperature (∼22°C) containing (mM): NaCl, 155; KCl, 5; glucose, 5; Mops, 10. In experiments where a Na⁺-free medium was required, N-methyl-D-glucamine (NMDG) was used as a replacement for sodium, unless otherwise stated. In experiments where a Cl⁻-free incubation medium was required, methylsulphate was used as the substitute anion. The pH value of all solutions was kept at 7.4 and the osmolarity checked using a Wescor 5100C vapour pressure osmometer. ‘MgCl₂ wash’ consisted of 107 mM MgCl₂ and 10 mM Mops (pH 7.4).

All reagents used were of the highest available grade of purity. The radioisotopes (³H-proline, ³H-glutamine, and ¹⁴C-glycine) were obtained from Amersham Life Science or from NEN Life Science Products. Chemicals were obtained either from Sigma or Merck. Unless otherwise stated, the stereoisomer form of all amino acids used in the experiments was the L-form.

Isolation of bovine chondrocytes

The isolation technique for chondrocytes was adapted from Kuettner et al. (1982). The cells released by this isolation have been extensively characterized and confirmed as articular chondrocytes on the basis of the expression of specific markers such as collagen II (for example, Kuettner et al. 1982; Zanetti et al. 1985). The forefeet of 12- to 36-month-old steers killed by abbattoir pithing were collected immediately after slaughter. The metacarpophalangeal joints were opened under near-aseptic conditions and shavings of hyaline cartilage (5 mm × 4 mm × 0.5 mm) removed from the outer two-thirds of the articular cartilage, such that contamination with bone cells or red blood cells could be avoided. Cartilage slices were pooled and maintained overnight at 37°C in DMEM containing 33 μg ml⁻¹ gentamicin and 33 μg ml⁻¹ penicillin plus 5 % (v/v) FCS.

After washing three times in DMEM, the cartilage was incubated in DMEM plus 1 % (w/v) protease type XXV and 5 % (v/v) FCS at 37°C for 1 h. The protease digests the proteoglycans leaving a partly unmasked network of collagen fibrils. The cartilage was then centrifuged (3000 g, 5 min), the supernatant discarded, and the resuspended pellet washed in fresh DMEM. After further washing, the cartilage was incubated in DMEM containing 900 units ml⁻¹ collagenase type I and 5 % (v/v) FCS for 2 h 45 min. Proteoglycan digestion by protease partly unmasks collagen fibrils, which renders them more susceptible to cleavage by collagenase and results in almost complete digestion of the remaining collagenous network in the intercellular matrix.

Typically, chondrocytes suspended in the digested matrix are surrounded by remnants of the cartilage. Individual cells were collected by passing the digest through a Nitex filter (250 μm pore size). Half of the isolated cell suspension was washed three times in the relevant ice-cold incubation medium and kept on ice whilst the cells were counted, and checked for cell viability using Trypan Blue dye exclusion. The other portion of the cells was washed three times in DMEM and maintained overnight at 37°C in DMEM containing 5 % FCS (v/v), 33 μg ml⁻¹ gentamicin and 33 μg ml⁻¹ penicillin. These were used the following day for a repeat experiment after washing in the incubation medium and checking of cell viability.

Cell counting was necessary for evaluation of cell viability and standardization of subsequent flux rates as per million cells. A 100 μl aliquot of the isolated chondrocyte suspension in incubation medium was added to 200 μl of 0.5 % (w/v) of the vital dye Trypan Blue. A sample was placed on an improved Neubauer haemocytometer stage and examined under a light microscope. The viability was always more than 95 %. A value less than this was considered too low for accurate flux studies, and the cells discarded.

Flux protocol

To measure amino acid uptake, the chondrocytes were incubated at 37°C in 0.4-1.0 ml of the required medium. The reaction was started by addition of the tracer amino acid. After incubation the cells were separated by centrifugation (10000 g, 10 s), washed in ice-cold isotonic MgCl₂ solution (four times), and processed for β-scintillation counting, as previously described (Ellory et al. 1981).

The flux was calculated from the mean of triplicate samples and expressed in mmol (10⁶ cells)⁻¹ min⁻¹. Results quoted are the means ± s.e.m. of independent fluxes unless otherwise indicated. The number of independent determinations (n) used to obtain the mean is also stated.

Reverse transcription-polymerase chain reaction (RT-PCR)

Total RNA was extracted from bovine chondrocytes and rat brain tissue homogenate using the guanidinium thiocyanate method (Chomczynski & Sacchi, 1987) and resuspended in nuclease-free water. cDNA was reverse transcribed from this RNA using the Superscript II reverse transcription system and oligo dT₁₅ primer according to the manufacturer's protocol (Life Technologies, UK). Primers for the PCR amplification were chosen from human β-actin mRNA sequence and consensus sequences within cloned GLYT-1 (all isoforms), GLYT-2 and PROT mRNA sequences available in the GenBank nucleotide database (The National Centre for Biotechnological Information, NIH, USA) using Primer 0.5 (Whitehead Institute, MIT, USA). The primer sequences used are given in Table 1. Genbank accession numbers for representative rat (transporter) and human (β-actin) coding sequences are given in the table, and the corresponding base location of these primers in these coding sequences stated.

Table 1.

Primers for PCR amplification

PCR primer	Sequence (5′ > 3′)	Accession no. of coding sequence	Base position of primer (5′ > 3′) in this sequence
GLYT-1 (forward)	CTGGAGGCTGTATGTGCTGA	M95413	726–745
GLYT-1 (reverse)	GATGACGAAGCCAGCATAGA	M95413	844–825
GLYT-2 (forward)	ACCAACATCTTGGAGGCAAC	L21672	247–266
GLYT-2 (reverse)	CTCACGTTCTCGTCCTCCTC	L21672	617–598
PROT (forward)	ACCAGACACCTCCAAACAGG	M88111	2000–2019
PROT (reverse)	GGTCCAGCTCTTCTCTCCCT	M88111	2278–2259
β-Actin (forward)	TTCAACTCCATCATGAAGTGTGACGTG	M10277	2595–2621
β-Actin (reverse)	CTAAGTCATAGTCCGCCTAGAAGCATT	M10277	3016–2990

PCR reactions were set up using 2 μl cDNA from each RT reaction, PCR buffer (1 mM Tris-HCl, 500 mM KCl, pH 8.3), 1.5 mM MgCl₂ (1.0 mM for GLYT-2 primers), 200 μM of each dNTP (Pharmacia Biotech, UK), 50 pmol of each paired primer, 1 % DMSO, 1.5 units Taq DNA polymerase (Boehringer Mannheim) in 50 μl volumes with nuclease-free H₂O. Reactions were heated to 95°C for 2 min followed by 40 cycles of denaturation at 95°C for 1 min, primer annealing at 58°C for 1 min, extension at 72°C for 2 min and then one final cycle of extension at 72°C for 5 min in a thermal cycler equipped with heated lid (Hybaid, UK).

For analysis of PCR products, 20 μl samples from the reactions were size-fractionated on 2 % NuSieve GTG agarose gels (Flowgen, UK) containing 0.5 μg ml⁻¹ ethidium bromide in 40 mM Tris acetate-1 mM EDTA, pH 8 at 50 V for 2 h.

RESULTS

Initial rates of uptake

The influx of each amino acid (extracellular concentration 0.5 mM) was linear for at least 8 min at 37°C. The mean slopes of the linear part of uptake with time for proline, glycine and glutamine are (× 10⁻⁸ mmol (10⁶ cells)⁻¹ min⁻¹): 3.26 ± 0.26, 5.6 ± 1.3 and 34.7 ± 10.9, respectively (n = 3). Subsequent flux experiments were performed with incubation times of 5 min.

For transport studies with radioactive amino acids, incorporation into protein or metabolism of the amino acid intracellularly is a possible problem, including conversion of the substrate to other metabolites, or carbon dioxide and water and subsequent efflux of the radioisotope. This is unlikely to be relevant in the present case for the following reasons. Firstly, the rate of metabolism in the chondrocyte is known to be very slow (Stockwell, 1979; Fassbender, 1987). Secondly, early studies by Byers (1983) had shown that no significant metabolic conversion of various neutral amino acids took place with up to 20 min incubation at 37°C in bovine chondrocytes. In addition, incorporation of the test amino acid into acid-precipitable protein was not evident after 5 min incubation as judged by a comparison of TCA (trichloroacetic acid)- and non-TCA-treated samples (results not shown). Furthermore, thin layer chromatography with butanol:acetic acid:water of carrier proline (Block et al. 1958) confirmed that 60-75 % of the radioactivity in the supernatant after TCA precipitation co-chromatographed with proline. Therefore possible aberrations in flux rates as a result of incorporation of the isotope into the protein can be discounted. Non-specific osmotic effects of competitive substrates were discounted by influx experiments performed in the presence of 20 mM sucrose: no change in uptake could be detected in hyperosmotic conditions (data not shown).

Proline transport

One major problem in the identification of amino acid transport systems is the lack of specific inhibitors. Only the non-metabolizable paradigm substrate for system A, α-methylaminoisobutyric acid (MeAIB), has proved to be specific enough in all cell types studied (Christensen et al. 1965; Christensen, 1989). Alanine has been shown previously to be transported by both systems A and ASC (Barker & Ellory, 1990a). These two substrates therefore could be used to show whether systems A and ASC were responsible for proline uptake.

Chondrocytes were incubated at 37°C for 5 min in the standard saline medium containing either 0.5 mM proline alone or proline with increasing concentrations of MeAIB from 0 mM to saturation, that is a level at which no more inhibition of proline uptake took place. Then, keeping the MeAIB concentration constant, the concentration of alanine was increased until maximal inhibition of proline transport had occurred. Influx of proline was investigated subsequently under identical conditions but in an Na⁺-free incubation medium (using an equimolar concentration of NMDG and retaining Cl⁻ concentration at 155 mM).

Neither MeAIB nor alanine had any effect on proline transport in the absence of Na⁺ (data not shown). However, in the presence of Na⁺, increasing amounts of MeAIB up to 3 mM progressively inhibited proline transport. Above this concentration the rate of proline uptake remained constant. Similarly, for alanine concentrations up to 1 mM, proline influx was inhibited further, until complete inhibition of proline transport was achieved.

Using MeAIB as an inhibitor of system A, at an extracellular proline concentration of 0.5 mM, 25.5 ± 2.3 % (n = 9) of total uptake was via system A and 40.7 ± 2.4 % (n = 9) was via system ASC. The remaining 33.8 ± 2.4 % (n = 9) of uptake was by Na⁺-independent routes.

Proline uptake via systems A and ASC was characterized kinetically by varying the concentrations of extracellular proline in the presence or absence of Na⁺ ions. If the rate of proline influx is determined in the standard saline medium, at varying concentrations of proline, but in the presence of 20 mM MeAIB (to inhibit transport via system A), it should then be possible to dissect the Na⁺-dependent component of proline uptake into systems A and ASC, both of which can be analysed kinetically.

Figure1A shows the results of a typical experiment. In the absence of Na⁺, proline uptake was linear with concentration. A more detailed analysis of the Na⁺-dependence data from the same experiment is shown in Fig. 1B where the rate of proline uptake via the two Na⁺-dependent components has been plotted. The data for each component could be fitted to the Michaelis-Menten equation and the kinetic constants, K_m and V_max, determined. The mean values of the kinetic constants from four separate determinations using MeAIB as an inhibitor of system A are shown in Table 2.

Figure 1. Concentration dependence of proline influx.

A, the concentration dependence of proline influx in standard buffered saline (^) and medium containing NMDG (▵). The concentration dependence was repeated in saline but in the presence of 20 mM MeAIB (•). In this and all subsequent flux experiments, each data point is the mean ± s.e.m. of triplicate samples. B, the concentration dependence of proline influx divided into the MeAIB-sensitive component (^) and MeAIB-insensitive, Na⁺-dependent component (•). Curves were derived from the same experiment shown in Fig. 1A. Concentration of MeAIB used was 20 mM. The computed kinetic parameters for the MeAIB-sensitive component in this single determination are: K_m = 1.96 ± 0.50 mM, V_max = (5.67 ± 0.74) × 10⁻⁸ mmol (10⁶ cells)⁻¹ min⁻¹. For the residual Na⁺-dependent component (MeAIB-insensitive): K_m = 0.24 ± 0.07 mM, V_max = (1.37 ± 0.08) × 10⁻⁸ mmol (10⁶ cells)⁻¹ min⁻¹.

Table 2.

Summary of the Na⁺-dependent transport systems for the three amino acids studied

Amino acid	Na+-dependent amino acid transport systems	Km (mM)	Vmax(× 10−8 mmol (106 cells) min−1)
Proline	System A	1.88 ± 0.25	4.01 ± 0.65
	System ASC	0.75 ± 0.30	1.16 ± 0.18
Glycine	System A	2.08 ± 1.07	10.4 ± 1.4
	System ASC	8.04 ± 0.54	15.5 ± 2.3
	System Gly
	Method A	0.93 ± 0.20	2.96 ± 0.91
	Method B	0.85 ± 0.16	4.54 ± 1.57
Glutamine	System A	0.29 ± 0.09	5.74 ± 0.93
	System ASC	0.32 ± 0.09	5.40 ± 0.35
	System N	0.47 ± 0.08	6.70 ± 0.68

The kinetic constants, K_m and V_max, were derived by computer fitting the rates of substrate uptake to the Michaelis–Menten equation. Results shown are means ± s.e.m. (n ≥ 3). System Gly was measured in two ways: method A measured the residual Na⁺-dependent MeAIB-and serine-insensitive component of glycine influx; method B measured the Cl⁻-dependent component of glycine influx.

Glycine transport

In human erythrocytes, five separate components of glycine transport have been identified. The predominant pathway was the high affinity Na⁺- and Cl⁻-dependent system Gly. In addition, the Na⁺-dependent and Cl⁻-independent system ASC, the Na⁺-independent system L, the SITS-sensitive Band 3, and a residual non-saturable, Na⁺-independent component were found to be present (Ellory et al. 1981). In the chondrocyte, the number of Na⁺-dependent systems able to transport glycine may therefore be greater than for proline transport.

The rate of uptake of 0.5 mM glycine was measured in the presence and absence of Na⁺ and with increasing concentrations of the competitors MeAIB and serine as exemplified in Fig. 2. The serine was added to inhibit glycine uptake via system ASC. It can be seen that neither MeAIB nor serine had any effect on glycine uptake in the absence of Na⁺. However, in the presence of Na⁺, a progressive increase in [MeAIB] inhibited glycine uptake. An increase in serine concentration also inhibited glycine transport but a significant residual Na⁺-dependent component (> 2 × 10⁻⁸ mmol (10⁶ cells)⁻¹ min⁻¹) remained uninhibited by either MeAIB or serine. These findings indicate that glycine is transported by systems A and ASC and also by a third Na⁺-dependent component. The fractional inhibition of glycine influx by MeAIB and serine is shown in Table 3. It can be seen that the residual component represents more than 40 % of total Na⁺-dependent glycine uptake.

Figure 2. The effect of increasing concentrations of MeAIB and MeAIB plus serine in the presence (^ and •) and absence (▵) of Na⁺ ions on glycine influx.

Extracellular glycine concentration = 0.5 mM. NMDG was used as the Na⁺ replacement.

Table 3.

Inhibitory effect of various model substrates for amino acid transport systems on glycine influx (extracellular glycine concentration 0.5 mM)

Inhibitor (mM)	Percentage inhibition
Na+-dependent influx of glycine (0.5 mM)
MeAIB (6 mM)	23 ± 7
MeAIB (6 mM) + serine (3 mM)	59 ± 7
MeAIB (6 mM) + serine (3 mM) + sarcosine (20 mM)	104 ± 4
Cl−-dependent influx of glycine (0.5 mM)
Sarcosine (0.2 mM)	29 ± 9
Sarcosine (20 mM)	72 ± 6

The concentration at which the inhibitors were present is shown in parentheses. Data shown are means ± s.e.m. (n ≥ 3).

One likely candidate for glycine transport in these experiments is system Gly. This system has been identified previously in other cell types by its requirement for both Na⁺ and Cl⁻ ions and its relatively high specificity for glycine and sarcosine only. Therefore it should be possible to use these two criteria as a means of testing whether this third component is system Gly. The Na⁺-dependent uptake of 0.5 mM glycine was followed in the presence of 6 mM MeAIB and 3 mM serine (to inhibit systems A and ASC) and also in the presence and absence of 20 mM sarcosine. It can be seen from Table 3 that sarcosine was able to inhibit all of the glycine transport via this residual component.

This suggests that the residual Na⁺-dependent component represented a single system, that of system Gly. To test this hypothesis further, glycine uptake was followed in the presence and absence of Cl⁻ ions (using MeSO₄⁻ as the anionic substitute) and in the presence and absence of 20 mM sarcosine. A Cl⁻-dependent glycine component was observed which was largely inhibited by excess sarcosine (Table 3).

Each of the three Na⁺-dependent systems responsible for glycine transport could be characterized kinetically in a similar manner to that for the proline-transporting systems. The uptake of glycine was measured at varying concentrations in the presence and absence of Na⁺ in order to obtain the Na⁺-dependent rate of uptake. This was then repeated but in the presence of 20 mM MeAIB or 20 mM MeAIB plus 5 mM serine. The remaining MeAIB-serine-insensitive component corresponds to glycine transport via system Gly. The data were then computer-fitted to the Michaelis-Menten equation and the kinetic constants obtained (Table 2).

It was also possible to characterize system Gly kinetically using a different criterion. Incubation of the cells in the presence and absence of Cl⁻ ions at varying concentrations of glycine between 0.25 and 3 mM gave the Cl⁻-dependent rate of glycine uptake (e.g. Fig. 3). The rate of Na⁺-dependent glycine uptake in the presence of excess MeAIB and serine, as described above, has also been plotted for comparison. It can be seen that an analysis of either the residual Na⁺-dependent component or the Cl⁻-dependent component of glycine uptake produces a similar concentration dependence. A kinetic comparison of K_m and V_max values for each component is given in Table 2.

Figure 3. The concentration dependence of the initial rate of Na⁺-dependent MeAIB- and serine-insensitive glycine (^) and Cl⁻-dependent glycine influx (•).

In this representative experiment, the K_m and V_max of the Na⁺-dependent component are 0.89 ± 0.18 mM and (4.64 ± 0.43) × 10⁻⁸ mmol (10⁶ cells)⁻¹ min⁻¹, respectively. The computed kinetic parameters for the Cl⁻-dependent component are K_m = 1.05 ± 0.26 mM and V_max = (4.26 ± 0.43) × 10⁻⁸ mmol (10⁶ cells)⁻¹ min⁻¹, respectively.

Glutamine transport

System A, ASC and N have all been shown to be capable of transporting glutamine in certain other cell types (Kilberg et al. 1980; Handlogten et al. 1981). However, the systems responsible in the chondrocyte have not been identified. Inhibition studies of glutamine uptake using excess MeAIB revealed a MeAIB-sensitive component corresponding to approximately 31 % of total uptake (at 0.5 mM; data not shown). Kilberg et al. (1979) have used cysteine as a system-specific substrate for transport system ASC in the hepatocyte and thereby allowed subsequent characterization of system N (Kilberg et al. 1980). However, our preliminary experiments in the chondrocyte revealed cysteine was not specific to system ASC but was able to inhibit glutamine uptake via systems A, ASC and N. We have therefore employed serine to inhibit glutamine transport by system ASC. A typical result for glutamine influx in the presence of 8 mM MeAIB with increasing concentrations of serine and then histidine (an inhibitor of system N) is shown in Fig. 4. Serine is able to inhibit system ASC with a relatively high degree of specificity (exemplified by the levelling-off of the inhibition curve). In addition, histidine is also able to inhibit glutamine transport further. However, there is a small Na⁺-dependent component insensitive to MeAIB, serine and histidine inhibition. This final component, after subsequent inhibition studies with a variety of amino acids, was found not to correspond to any known transport system. At an extracellular glutamine concentration of 0.5 mM, approximately 28 % of total uptake was by system ASC and 22 % by system N. There was a slight inhibitory effect of serine and histidine (a substrate of system L) on Na⁺-independent glutamine uptake. This was unlike the action of serine on proline or glycine transport (e.g. see Fig. 2) and may suggest the presence of a Na⁺-independent amino acid pathway which is able to transport glutamine.

Figure 4. The effect of increasing concentrations of serine and serine plus histidine in the presence of (^) and absence (•) of Na⁺ ions on glutamine influx.

Extracellular glutamine concentration = 0.5 mM. MeAIB was present at 8 mM in the incubation medium to inhibit system A. NMDG was used as the Na⁺ replacement.

Experiments to measure the initial rates of uptake (described earlier) revealed that the rate of glutamine transport was much higher than that for proline or glycine at the same extracellular substrate concentration. Kinetic characterization of the Na⁺-dependent transport systems which carry glutamine revealed systems A and ASC to have a high affinity for glutamine relative to that for proline and glycine. In addition, system N was shown to be an avid transporter of glutamine (Table 2). The kinetic values were obtained by measuring the concentration dependence (in the range 0.25-2.0 mM) of the rate of Na⁺-dependent glutamine influx in the absence and then the presence of (a) 8 mM MeAIB; (b) 8 mM MeAIB plus 2 mM serine; and (c) 8 mM MeAIB, 2 mM serine and 5 mM histidine.

Finally, excess D- and L-glutamine were compared as inhibitors of Na⁺-dependent glutamine uptake. Ninety per cent of Na⁺-dependent glutamine uptake (0.5 mM extracellular concentration) was inhibited by a 20-fold excess of the L-form, compared with a 10 % inhibition by D-glutamine (n = 3). It can be seen that the L-form is much more effective than the D-form, in agreement with the known stereoselectivity of systems A, ASC and N, but in contrast with Na⁺-independent system L.

Detection of amino acid transporter mRNAs in isolated chondrocytes

Molecular evidence for the expression of GLYT, a component of system Gly, in chondrocytes is given in Fig. 5, which shows the amplification products from bovine chondrocyte and rat brain mRNA. Products of the predicted sizes for GLYT-1, GLYT-2 and PROT (439, 371 and 279 base pairs (bp), respectively) were detected in the positive control rat brain RT-PCR reactions, confirming the specificity of the primers used. However, of these three only a GLYT-1 product could be detected in the chondrocyte RT-PCR. No products were obtained when reverse transcriptase was omitted from the RT reaction. When the product was sequenced and compared with known sequences in the GenBank database, greater than 80 % homology with the cloned bovine retinal glycine transporter, GLYT-1 was found. Admittedly, it is possible that the bovine PROT cannot be amplified using human primers, although the rat brain samples serve as a cross-species control. The amplification product of β-actin mRNA (predicted size of 310 bp) served as a positive control for the RT-PCR reactions

Figure 5. Detection of GLYT-1, GLYT-2, PROT and β-actin mRNA in isolated bovine chondrocytes and positive control rat brain tissue by RT-PCR.

Middle lane, molecular weight markers (100 bp ladder).

Initial rates of uptake

An example of a time course of an uptake experiment for leucine and tryptophan is shown in Fig. 6. The influx of each amino acid (extracellular concentrations: leucine, 1.0 mM; tryptophan, 0.4 mM) was linear for at least 8 min at 37°C. The mean slopes of the linear component of uptake with time for leucine and tryptophan are (× 10⁻⁷ mmol (10⁶ cells)⁻¹ min⁻¹) 2.13 ± 0.46 and 1.54 ± 0.08, respectively (n = 3). Subsequent flux experiments were performed with incubation times of 5 min.

Figure 6. Influx of leucine (^) and tryptophan (•) over time in standard buffered (pH 7.4) saline medium at 37 °C.

Extracellular leucine concentration = 1.0 mM; extracellular tryptophan concentration = 0.4 mM. The first four data points fitted a linear regression with a slope of (1.67 ± 0.02) × 10⁻⁷ mmol (10⁶ cells)⁻¹ min⁻¹ for leucine uptake and (1.62 ± 0.05) × 10⁻⁷ mmol (10⁶ cells)⁻¹ min⁻¹ for tryptophan uptake.

Leucine transport

In order to identify the system(s) responsible for leucine transport, inhibition experiments were performed on leucine influx. Inhibitors tested consisted of a variety of amino acids and amino acid analogues, including 2-aminobicyclo (2,2,1)heptane-2-carboxylic acid (BCH), a specific inhibitor of system L in other mammalian cell types (Christensen et al. 1969).

Table 4 shows the mean percentage inhibition from three experiments on leucine uptake (extracellular concentration, 0.2 mM) in the presence of high concentrations of selected amino acids. MeAIB, proline and glycine all had a negligible effect on leucine transport in the absence of Na⁺. Glutamine was able to inhibit about 15 % of leucine uptake, and histidine and BCH were also able to produce significant inhibition.

Table 4.

Na⁺−independent leucine uptake (0.2 mM) in the presence of various other amino acids or analogues (10 mM) expressed as a percentage of leucine uptake in the absence of any inhibitor

Inhibitor (10 mM)	Uptake (% control)
Control	100
MeAIB	97 ± 2
Proline	94 ± 1
Glycine	93 ± 1
Glutamine	85 ± 3
BCH	20 ± 1*
Histidine	30 ± 5*

Values are means ± s.e.m. (n = 3).

^*Values statistically different from the control at the 99 % level (P < 0.01).

Further inhibition experiments on leucine uptake were carried out at varying concentrations of BCH and in the presence or absence of Na⁺. Figure 7 illustrates one of three similar curves demonstrating progressive inhibition. In the absence of Na⁺, 1 mM leucine uptake was shown to decrease with increasing BCH until near-maximal inhibition occurred at 8 mM BCH. An identical inhibition curve was revealed in the presence of Na⁺ thereby confirming BCH as specific for inhibiting the Na⁺-independent leucine system (data points at 0 and 8 mM BCH only have been included for the sake of clarity). One hundred per cent inhibition of total leucine uptake was not achieved (presumably because a proportion of leucine uptake occurs by diffusion).

Figure 7. Inhibition of leucine influx by increasing concentrations of BCH (2-aminobicyclo(2,2,1)- heptane-2-carboxylic acid).

Extracellular leucine concentration = 1.0 mM. Leucine influx was determined at concentrations of BCH of 0-8 mM in the presence (•) and absence (^) of Na⁺.

Leucine uptake via this BHC-sensitive system was characterized kinetically by measuring the rate of leucine uptake over the concentration range 0.1-0.8 mM in a Na⁺-free medium and in the presence and absence of 10 mM BCH. Figure 8 represents one of three such experiments in which the rate of leucine uptake under the two conditions has been plotted (mean of triplicate experiments). The difference between the rate of uptake in the absence of BCH and in its presence has also been plotted as the BCH-sensitive component. The latter component can be fitted to the Michaelis-Menten equation whereas the BCH-insensitive component is non-saturable and fits a simple linear regression. The mean K_m and V_max values of this system for leucine are 125 ± 27 μM and (1.49 ± 0.26) × 10⁷ mmol (10⁶ cells)⁻¹ min⁻¹, respectively (n = 3).

Figure 8. The concentration dependence of leucine influx in the absence (^) and presence (•) of 10 mM BCH.

Incubation medium was Mops-buffered NMDG medium. Subtraction of influx rate in the presence of BCH from the rate in the absence of BCH gives the BCH-sensitive component of leucine influx (▵). The computed kinetic parameters for the BCH-sensitive component in this single determination are: K_m = 123 ± 10 mM, V_max = (1.71 ± 0.04) × 10⁻⁷ mmol (10⁶ cells)⁻¹ min⁻¹. The BCH-insensitive component fitted a linear regression and had a slope of (0.8 ± 0.1) × 10⁻⁷ mmol (10⁶ cells)⁻¹ min⁻¹ mM⁻¹.

System L, in contrast to system N, has been shown to be only partially stereospecific in other cell types (Ellory, 1987). To verify further that leucine uptake is by a system L-like transporter in the chondrocytes, experiments were performed to determine the effect of D-leucine on L-leucine uptake. A fiftyfold excess of D-leucine was found to inhibit L-leucine uptake by 72 ± 3.7 % (n = 3), a value comparable to inhibition by BCH on L-leucine influx (see Table 2). Therefore system L in the chondrocyte can also be characterized as being poorly stereospecific.

Tryptophan transport

A comparison of tryptophan uptake (0.4 mM extracellular concentration) in buffered saline medium with uptake in Na⁺-free medium revealed almost identical rates of influx ((1.41 ± 0.02) × 10⁻⁷ mmol (10⁶ cells)⁻¹ min⁻¹ in Mops-buffered saline medium, (1.45 ± 0.04) × 10⁻⁷ mmol (10⁶ cells⁻¹) min⁻¹ in Mops-buffered NMDG medium, n = 3). Tryptophan uptake in chondrocytes is limited therefore to Na⁺-independent routes. To help identify which Na⁺-independent system(s) are used for tryptophan transport, inhibition studies which involved adding excess leucine, BCH, lysine and phloretin were carried out. The first two were used because they are preferred substrates of system L, lysine because system y⁺ has a high affinity for this substrate, and phloretin because of its known inhibitory action on tryptophan transport in the red blood cell (Rosenberg et al. 1980; Rosenberg 1981).

Table 5 illustrates the effects of these compounds on tryptophan uptake. Leucine, BCH and phloretin are able to inhibit uptake more than 50 %, but BCH and, especially, leucine blocked tryptophan transport by more than 70 % (leucine inhibited more than 80 % of uptake). However, lysine had no inhibitory effect on tryptophan transport, thereby eliminating system y⁺ as the high affinity Na⁺-independent system. Therefore the route by which tryptophan is taken up is system L-like.

Table 5.

Effect of leucine, BCH, phloretin (3-[4-hydroxyphenyl]-1-[2,4,6-tri-hydroxyphenyl]-1-propanone) and lysine on tryptophan influx

Inhibitor (mM)	Uptake (% control)
Control	100
Leucine (5 mM)	17 ± 1*
BCH (5 mM)	29 ± 1*
Phloretin (1 mM)	36 ± 1*
Lysine (10 mM)	93 ± 2

Cells were incubated in Na⁺-free medium, where the extracellular concentration of tryptophan was 0.4 mM except in the case of inhibition by lysine where the tryptophan concentration was 0.1 mM. The concentration of inhibitors is given in the table. Results have been expressed as a percentage of the rate of tryptophan influx in the absence of any inhibitors (control). Values are means ± s.e.m.

^*P < 0.001 (n = 3). Inhibition by lysine had a P value more than 0.05 (n = 3).

In the hepatocyte, both BCH and leucine were able to inhibit tryptophan uptake on system T by 50 % (Salter et al. 1986). A comparison of the inhibitor experiments from the chondrocyte with those for the hepatocyte suggests that system T may be absent in the former. If further inhibition can be produced by the addition of tyrosine (a preferred substrate of system T) in the presence of leucine, then this would support the presence of system T. However, the addition of 5 mM tyrosine in the presence of 10 mM leucine was found to decrease the rate of tryptophan influx from (0.38 ± 0.04) to (0.32 ± 0.03) × 10⁻⁷ mmol (10⁶ cells)⁻¹ min⁻¹ (P = 0.13, n = 3) and did not therefore have any additional inhibitory effect on tryptophan transport.

In conclusion, system T is probably absent from the chondrocyte. This is supported by the observation that in the erythrocyte, where system T is known to operate, leucine produced only slight inhibition of tryptophan uptake (Vadgama & Christensen, 1985). In addition, in the hepatocyte, where system T is also present, the addition of tyrosine in the presence of BCH did have a further inhibitory effect on tryptophan transport (Salter et al. 1986).

Leucine transport studies have shown that a high affinity L-like system is present and therefore it is possible for the same system to be used for tryptophan uptake. This can be investigated further by measuring the affinity of the system for tryptophan.

Figure 9, one of three typical experiments, shows how the rate of tryptophan uptake increases with tryptophan concentration in an approximately hyperbolic manner. In the presence of excess leucine, tryptophan transport is much reduced and shows a linear dependence on tryptophan concentration. The slope of this leucine-insensitive component was (1.03 ± 0.35) × 10⁻⁷ mmol (10⁶ cells)⁻¹ min⁻¹ mM⁻¹ (n = 3), and probably represents passive diffusion.

Figure 9. The concentration dependence of tryptophan influx in absence (^) and presence (•) of 10 mM leucine.

Incubation medium was Mops-buffered NMDG solution. Subtraction of influx rate in the presence of leucine from the rate in the absence of leucine gives the leucine-sensitive component of tryptophan influx (♦). The computed kinetic parameters for the leucine-sensitive component in this single determination are K_m = 58 ± 4 mM, V_max = (1.73 ± 0.06) × 10⁻⁷ mmol (10⁶ cells)⁻¹ min⁻¹. The leucine-insensitive component fitted a linear regression and had a slope of (1.5 ± 0.2) × 10⁻⁷ mmol (10⁶ cells)⁻¹ min⁻¹ mM⁻¹.

The leucine-sensitive component could be fitted to the Michaelis-Menten equation and the kinetic parameters determined. The mean K_m of this system for tryptophan was 35.6 ± 11.4 μM and the V_max was (1.92 ± 0.54) × 10⁻⁷ mmol (10⁶ cells)⁻¹ min⁻¹ (n = 3). This indicates that the transport system has an extremely high affinity for tryptophan and prefers tryptophan over leucine.

DISCUSSION

The present aim has been the identification of the transport systems responsible for uptake of the amino acids proline, glycine, glutamine, leucine and tryptophan into bovine chondrocytes.

System A was identified by its Na⁺ dependence and the inhibition of substrate influx using the substrate, MeAIB, although it must be recognized that even this inhibitor is not entirely specific (also being a substrate for a H⁺-driven amino acid transport system described in epithelial tissues, for example by Thwaites et al. 1993). All the amino acids under investigation were shown to be transported by system A. The apparent K_m values correspond well to investigations in other cell types. For instance, in the chondrocyte, system A had a K_m of 1.9 mM (see Table 2) for proline compared with a K_m of 1.3 mM in the lymphocyte (Segal et al. 1983).

In the absence of any system-specific inhibitors of system ASC, its presence in the chondrocyte was harder to assign. However, the demonstration of a MeAIB-insensitive, Na⁺-dependent component of proline transport which was completely inhibited by alanine provided evidence for the system. Similar inhibition studies on glycine and glutamine but by using serine instead of alanine added further proof.

System Gly was thought to be present in the chondrocyte when it was shown that not all Na⁺-dependent glycine influx could be inhibited by excess MeAIB and serine. More definitive evidence for its presence came from adding MeAIB, serine and excess sarcosine (a substrate of system Gly) to the extracellular medium and then measuring glycine influx. In this case 100 % inhibition of Na⁺-dependent glycine transport occurred. Further evidence was obtained from showing that a component of glycine transport required Cl⁻ ions. This Na⁺ and Cl⁻ dependence of system Gly enabled kinetic characterization by using two approaches: firstly, a study of the residual MeAIB- and serine-insensitive, Na⁺-dependent component of glycine influx; secondly, a kinetic study of Cl⁻-dependent glycine influx. Either method gave similar affinity constant values for glycine transport. The GLYT family of glycine transporters display Na⁺- and Cl⁻-dependent glycine transport activity that can be blocked by sarcosine but not by MeAIB or L-alanine and are therefore thought to be related to system Gly (Guastella et al. 1992; Malandro & Kilberg, 1996). In this study we have provided evidence that mRNA for at least one isoform, GLYT-1, is transcribed by chondrocytes and hence, if functionally expressed, is likely to contribute to the sarcosine-sensitive, non-system-A component of glycine transport in these cells. Furthermore, the inhibition by sarcosine of glycine uptake which we report here (Table 3) is consistent with GLYT-1 mediating glycine transport in chondrocytes (Liu et al. 1994). The GLYT family of transporters have been reported to have lower K_m values than the affinity presently reported, which may reflect the existence of a novel isoform, since this family has been shown to comprise several proteins originating from the same gene (Kim et al. 1994). In the same experiment we showed that mRNA for the GLYT-related proline transporter, PROT, was not expressed in chondrocytes; this protein is also sensitive to sarcosine. The absence of PROT was not surprising as expression of this protein is thought to be restricted to tissues involved in excitatory neurotransmission (Fremeau et al. 1992). A second, Na⁺-independent, H⁺-dependent, proline transporter (which may also perform significant transport of other amino acids, including alanine and glycine) has been identified in cultured intestinal cells (Thwaites et al. 1993) and so it will be interesting to determine whether this protein is expressed in chondrocytes once its nucleotide sequence is known.

System N, like system Gly, has a narrow substrate specificity and tissue distribution. This system was identified in the present investigation by the ability of histidine to inhibit the Na⁺-dependent glutamate influx whilst systems A and ASC were inhibited. System N was shown to have a K_m value of 0.47 mM for glutamine, a value not too dissimilar from that of 1.1 mM in the hepatocyte (Kilberg et al. 1980) or 0.28 mM in murine P388 leukaemia cells (Lazarus & Panasci, 1986).

System L has been found to have a ubiquitous tissue distribution. Classically, this system has been defined as the component of Na⁺-independent amino acid uptake that is inhibitable by BCH. Unfortunately, this characterization was premature with system T and systems L₁ and L₂ now known to be inhibited by this bicyclic analogue as well. The identification of system L in the chondrocyte was achieved by using BCH to inhibit leucine influx. The first indication that this system was not the classical system L was when kinetic studies revealed a very high affinity system (micromolar range) for leucine. Curiously, tryptophan was found to be transported even more avidly than leucine by this system. However, no inhibition of tryptophan influx by tyrosine occurred, suggesting that system T was not present. An investigation of the stereospecificity of system L showed it to be not unlike that of other tissues, with D-leucine competing with L-leucine for transport on the carrier. In conclusion, the system L identified in the chondrocyte has many properties similar to the ‘classical’ system L with the exception of having a high affinity for its substrates. It seems increasingly probable, given that high affinity L systems have also been identified in the hepatocyte, lymphocyte and at the blood-brain barrier (Weissbach et al. 1982; Segal et al. 1983; Christensen, 1989), that this high affinity system is more widely distributed than first thought. It may in fact be the case that this system is the housekeeping transporter and the low affinity system L described classically is an isoform or variant with more specific tissue distribution.

The rates of uptake of tryptophan and leucine in the chondrocyte are high. This can be explained by the relatively high substrate affinities of chondrocyte system L for leucine (K_m = 125 ± 25 μM) and tryptophan (K_m = 36 ± 11 μM). It is presumably no accident that this high affinity system is primarily responsible for the transport of essential amino acids that are present at low extracellular levels (132 μM leucine and 45 μM tryptophan in adult human plasma; taken from Geigy Scientific Tables, 1984). If amino acid concentrations in the extracellular matrix are found to be lower than plasma levels, the high affinity of the chondrocyte system would be crucial for the chondrocyte in ‘scavenging’ necessary amino acids. This would not only provide essential amino acids for protein biosynthesis, but could also load the cells with common amino acids and aid the uptake of other amino acids by trans-stimulation of exchange. Finally, it is curious that certain important amino acids such as glutamine, proline and glycine are acquired by the chondrocyte via Na⁺-dependent routes whilst other rarer amino acids are transported by a Na⁺-independent high affinity facilitated diffusion transport system.