Effects of Platelet-Rich Plasma Composition on Anabolic and Catabolic Activities in Equine Cartilage and Meniscal Explants

Abstract

Objective:

To evaluate the effects of single- and double-spin preparations of platelet-rich plasma (PRP) on anabolic and catabolic activities of cartilage and meniscal explants in vitro.

Methods:

Single- and double-spin PRP was prepared using laboratory processing or commercial kits. The cellular contents were quantified, and each PRP was mixed in equal quantities with cell culture medium and added to cartilage or meniscus explant cultures, with or without interleukin 1 β (IL-1β). Extracellular matrix synthesis was quantified over 24 hours via ³⁵S-sulfate and ³H-proline incorporation, while gene expression of catabolic enzymes was evaluated using real-time PCR.

Results:

The platelet concentration in single-spin laboratory PRP was 59% higher than blood. Platelet and white blood cell concentrations in single-spin laboratory and kit PRP were not significantly different, while the double-spin kit resulted in approximately 2.5-fold higher platelet and approximately 400-fold higher white blood cell concentrations. In cartilage cultures without IL-1β, radiolabel incorporation in single-spin PRP cultures was significantly higher than in double-spin cultures. Similar results were obtained for ³⁵S-sulfate incorporation in meniscus cultures without IL-1β. In IL-1β, radiolabel incorporation was largely similar among all PRPs. After 24 hours of culture, ADAMTS-4 gene expression in cartilage was lowest for single-spin PRP, while expression in the double-spin kit was not significantly different from double-spin laboratory PRP in which platelets were concentrated 6-fold.

Conclusions

This study suggests that single-spin PRP preparations may be the most advantageous for intra-articular applications and that double-spin systems should be considered with caution.

Introduction

Platelets represent an accessible and inexpensive source of growth factors that have recently received considerable attention for the treatment of injured or damaged musculoskeletal tissues. ¹ Enthusiasm for the therapeutic potential of platelets is based on the α-granules that contain growth factors such as transforming growth factor β (TGFβ) and platelet-derived growth factor (PDGF). ² The growth factor concentration in platelets, coupled with the platelet concentration in blood, is sufficient to yield nanograms of growth factor per milliliter of blood without increasing the native platelet concentration. ³

Platelet growth factors are typically administered in plasma, the combination of which is known as platelet-rich plasma (PRP). PRP is prepared from whole blood using centrifugation steps that eliminate the majority of the red blood cells and concentrate platelets based on the specific gravities of each cell type. ⁴ In some cases, PRP protocols result in a significant white blood cell content. ⁵ While extensive efforts are underway to understand the therapeutic benefit of PRP for tendon, ligament, and bone, ¹ relatively few studies have evaluated the effect of PRP on articular cartilage and the meniscus. In a laboratory study, PRP was shown to stimulate chondrocyte extracellular matrix (ECM) synthesis. ⁶ When administered via intra-articular injection in ovine subjects, PRP increased cartilage healing in a microfracture treatment model for chondral defects. ⁷ In addition, PRP delivered in a gelatin carrier stimulated superior healing relative to platelet-poor plasma in a meniscal defect model. ⁸ In human cases of chronic osteoarthritis, intra-articular injections of PRP resulted in improvements in clinical scores. ⁹ Based on these data, PRP may hold promise for stimulating tissue growth in degenerative or damaged cartilage and meniscus.

An important factor in characterizing the potential of PRP to heal joint tissues is to determine the dose of platelets that best stimulates cartilage or meniscal growth. PRP containing less than a 2-fold increase in platelet concentration relative to blood may be obtained by a single-spin protocol in which the red blood cells are pelleted and the entire plasma supernatant is collected. ¹⁰ To further concentrate the platelets, a second higher force spin may centrifuge the platelets into a pellet, after which the platelets are resuspended in a reduced plasma volume. ⁴ The platelet concentration in PRP can be at least 10-fold greater than that in whole blood, which results in TGFβ1 concentrations in excess of 50 ng/mL.^11,12 Such TGFβ1 concentrations may be favorable for stimulating ECM synthesis in articular joint tissues, as laboratory studies have demonstrated increasing proteoglycan synthesis with increasing TGFβ1 concentration up to 10 or 100 ng/mL in cartilage ¹³ or meniscal ¹⁴ explants, respectively. Based on these data, we hypothesized that double-spin PRP preparations will stimulate higher ECM synthesis without a concomitant increase in catabolic activities in cartilage and meniscal explants than single-spin protocols in which the platelets are minimally concentrated. We tested this hypothesis using an explant laboratory model that used tissues harvested from adult horses, an ideal species for modeling human cartilage resurfacing. ¹⁵

Materials and Methods

Equine Tissues

Tissues were obtained from 5 healthy 2- to 5-year-old horses. Prior to sedation for euthanasia, 50 mL of blood was drawn into 60-mL syringes preloaded with 5 mL of anticoagulant citrate dextrose using a 14-gauge medical-grade catheter. Following euthanasia, articular cartilage and menisci were retrieved from the femorotibial joint, equivalent to the human knee. Each femorotibial joint was grossly normal, and tissue samples were harvested within 3 hours of euthanasia. Full-thickness (approximately 1-2 mm) sections of articular cartilage were removed from the main load-bearing region of one medial femoral condyle and stored in sterile phosphate-buffered saline (PBS). Medial menisci were removed from each subject joint and stored in sterile PBS. Cartilage sections were divided into approximately 5-mm by 5-mm explants and placed in wells of a 48-well plate in low glucose DMEM (Invitrogen, Carlsbad, CA). Approximately 3-mm-thick meniscal explants were harvested from the femoral articulating surface of the avascular portion (inner two thirds) of the medial menisci. As performed for cartilage, approximately 50-mg samples were cut from the explants and stored in 48-well plates in low glucose DMEM. Once all explants were distributed into wells (approximately 1 hour), the medium was replaced with PRP medium made as described below.

Preparation of PRP and Culture Medium

PRP was processed within 2 hours of blood collection.

Commercial Kit Preparations of PRP

Single spin

PRP was prepared using Arthrex ACP (Naples, FL) by centrifuging the kit at 160g for 5 minutes and collecting the entire plasma supernatant above the red blood cell pellet. Hereafter, this PRP is referred to as “single-spin kit”.

Double spin

PRP from Harvest Technologies SmartPrep 2 (Plymouth, MA) was prepared according to the manufacturer’s protocol. Hereafter, this PRP is referred to as “double-spin kit”.

Laboratory Preparations of PRP

Blood was transferred to centrifuge tubes and spun at 200g for 18 minutes. The plasma was harvested down to the interface between the pelleted red blood cells. A portion of this PRP, referred to hereafter as “single-spin laboratory”, was set aside for experimentation. The remaining PRP was centrifuged at 1,000g for 15 minutes, which pelleted approximately 98% of the platelets (data not shown) and resulted in a platelet-poor plasma (PPP) supernatant. PPP was removed from each tube and was added back to the centrifuged platelets in reduced volumes that resulted in 3x, 6x, and 9x concentrations of platelets relative to single-spin laboratory PRP. Hereafter, these PRP preparations are referred to as “double-spin 3x”, “double-spin 6x”, and “double-spin 9x”, respectively. The platelets were resuspended by gently pipetting with a 5-mL pipette. All PRP preparations appeared yellow in color and absent of red discoloration that would indicate hemolysis.

Cell Counts

The platelet concentration in whole blood was determined via automated counts performed in the Diagnostic Laboratory at the Orthopaedic Research Center, Colorado State University. The concentration of platelets in PRP was determined by diluting 1:20 in PBS and manually counting on a hemocytometer. We confirmed that manual counts were not significantly different than automated counts (data not shown). White blood cells were counted using a hemocytometer. White blood cell differential counts were not performed.

Culture Medium

PRP was mixed with equal volumes of low glucose DMEM (containing 1.8 mM calcium chloride) supplemented with 10 mM HEPES, 0.1 mM nonessential amino acids (Sigma Chemical, St. Louis, MO), 20 µg/mL ascorbate-2-phosphate (Wako Chemicals, Richmond, VA), 100 U/mL penicillin G, 100 ug/mL streptomycin, and 0.25 mg/mL amphotericin B. For each tissue and medium formulation, half of the cultures received 10 ng/mL of recombinant human IL-1β (R&D Systems, Minneapolis, MN) resuspended in PBS plus 0.1% bovine serum albumin (Sigma Chemical) to produce a proinflammatory environment, as previously performed for equine cartilage. ¹⁶

Evaluation of ECM Synthesis

For each condition and horse, one explant was evaluated for ³⁵ S-sulfate and ³ H-proline incorporation as measures of proteoglycan and protein synthesis, respectively. Radiolabel medium was prepared by adding 5 uCi/mL of each isotope in volumes that were less than 1% of the total. After 24 hours of incubation in a cell culture environment, the explants were washed 5 times over 90 minutes in PBS plus 1 mM unlabeled proline and sulfate. Each explant was transferred to 1 mL of proteinase K–TRIS HCl solution and digested overnight at 60 °C. From the digests, radiolabel incorporation was measured via scintillation counting. These data were normalized to total DNA, which was evaluated by mixing aliquots of digest with Hoescht dye and quantifying against calf thymus DNA standards on a spectrofluorometer. ¹⁷

Gene Expression of Catabolic Enzymes

After 24 hours of culture, 4 additional explants per condition and horse were pooled, snap frozen and pulverized in liquid nitrogen, stabilized in TRIzol reagent (Invitrogen), and then homogenized for 50 seconds on ice. The samples were centrifuged for 10 minutes at 10,000g, and the supernatant was saved. Nucleic acids were extracted from the supernatant by adding 20% chloroform (volume/volume) and then centrifuging for 15 minutes at 12,000g. This step was repeated 3 to 5 times to ensure separation of the nucleic acids from the residual ECM. ¹⁸ The nucleic acids were precipitated by adding a 5x volume of isopropanol and then centrifuging for 10 minutes at 10,000g. The pellet was resuspended in nuclease-free water, and absorbance was measured at 260/280 nm to determine concentration and purity. For samples that resulted in 260/280 ratios less than 1.7, the RNA was further purified using the RNeasy Mini Kit (Qiagen, Chatsworth, CA) according to the manufacturer’s instructions. RNA was reverse transcribed to cDNA using 50 µM random hexamers (Superscript III cDNA synthesis kit, Invitrogen). cDNA solution and TAQMan Gene Expression Master Mix (Applied Biosystems, Foster City, CA) were mixed with primers/probes for the genes of interest, and expression levels were determined using semiquantitative real-time PCR conducted on an ABI Prism 7000 Sequence Detection System (Applied Biosystems). Primers/probes for ADAMTS-4, ADAMTS-5, MMP1, and GAPDH were designed and validated at our institution (ADAMTS-4: forward-TGTGATCGTGTCATTGGCTCC, reverse-TGTTTGCTGCAGCTAGAACCATC, probe-AGTTTGACAAGTGCATGGTGTGCGGT; ADAMTS-5: forward-AAGGTGACTGATGGGACCGAATGT, reverse-TTTGAGCCAATGATGCCGTCACAG, probe-AGGCCATACAGTAATTCCGTCTGCGT; MMP1: forward-ACTGCCAAATGGACTTCAAGCTGC, reverse-TCTTCACAGTGCTAGGAAAGCCG, probe-CAGGATGTGCTCTACGGATACCCCAAGGAC; GAPDH: forward-AAGTGGATATTGTCGCCATCAAT, reverse-AACTTGCCATGGGTGGAATC, probe-TGACCTCAACTACATGGTCTACATGTTTCA). Primers and probes for MMP13 were purchased from the Lucy Whitter Molecular Core Facility (Davis, CA). Gene expression was normalized to the housekeeping gene GAPDH. For each gene, relative expression was calculated by calibrating to the average expression in single-spin laboratory cultures. ¹⁹

Evaluation of Growth Factors Released into the Medium

TGFβ1 released to the culture medium in IL-1β–free cultures was evaluated using a commercial kit (R&D Systems, Minneapolis, MN) that has been validated for equine proteins. ²⁰ Prior to the start of culture, baseline TGFβ1 concentrations were evaluated for medium supplemented with single-spin laboratory, single-spin kit, and double-spin kit PRP. Following 24 hours of culture, media from IL-1β–free cartilage and meniscus cultures were evaluated for TGFβ1 content.

Data Analysis

The cellular content of single-spin laboratory PRP relative to blood was compared for 4 donor horses using a paired t test. The analysis of cartilage and meniscal explants was repeated for blood and tissue samples from 5 donor horses, with blood and tissues paired in an autologous fashion. TGFβ1 concentration and gene expression data were log transformed to meet the assumption of normality as determined by residual plots. All data were analyzed using a mixed-model analysis of variance (Glimmix procedure, SAS 9.2, SAS Institute, Carey, NC), with the donor animal used as a random effect. Individual comparisons were made using a least-squares means procedure. For cell and TGFβ1 concentration outcomes, PRP preparation was considered as the main effect. For radiolabel and gene expression outcomes, PRP preparation and IL-1β were considered main effects as well as their interactions. Individual comparisons of main effects or interactions were indicated based on a protected F test. P values less than 0.05 were considered significant. Data are reported as mean ± standard error of the mean.

Results

Cellular Content of PRP Preparations

Preliminary tests were performed to determine the extent to which platelets in single-spin laboratory PRP were concentrated relative to whole blood. The platelet concentration in blood was 172 ± 12 thousand/µL, which is consistent with a previous report for equine blood (mean = 165 thousand/µL ¹¹ ). Following centrifugation, the platelet concentration was 275 ± 27 thousand/µL, representing a 59% increase over blood (P < 0.01). The packed cell volume, calculated as a percentage of the volume of the red blood cell pellet following centrifugation at 1,000g for 10 minutes relative to the total volume, was 43% ± 1%.

For PRP preparations used for cartilage and meniscus cultures, the concentration of platelets and white blood cells was determined for single-spin laboratory, single-spin kit, and double-spin kit PRP. The mean concentration of platelets in single-spin laboratory PRP (317 ± 22 thousand/µL) ( Fig. 1A ) was not significantly different from single-spin kit PRP (276 ± 8 thousand/µL) (P = 0.20). Double-spin kit PRP contained 2.3- and 2.6-fold higher platelet concentrations (725 ± 95 thousand/µL) relative to single-spin laboratory and single-spin kit PRP, respectively. White blood cell counts in single-spin laboratory PRP (0.04 ± 0.001 thousand/µL) were not significantly different from single-spin kit PRP (0.03 ± 0.01 thousand/µL) (P = 0.98) ( Fig. 1B ). Double-spin kit PRP contained at least a 400-fold higher concentration of white blood cells (14.8 ± 3.0 thousand/µL) relative to single-spin laboratory and single-spin kit PRP.

Figure 1.

Cell content of laboratory, ACP, and SmartPrep preparations of PRP. (A) Platelet concentration. For each assay, different letters indicate significant differences (P < 0.05). (B) White blood cell concentration. For each assay, different letters indicate significant differences (P < 0.05)

Concentration of TGFβ1 in Culture Medium

Baseline TGFβ1 in Uncultured Medium

Prior to adding to explant cultures, samples of single-spin laboratory, single-spin kit, and double-spin kit medium were centrifuged to remove the cells, and the baseline TGFβ1 content was measured. The concentrations of TGFβ1 in uncultured single-spin laboratory (365 ± 130 pg/mL, abbreviated as “Base” in Fig. 2 ), single-spin kit (463 ± 241 pg/mL, data not shown), and double-spin kit (577 ± 93 pg/mL, data not shown) media were not significantly different (P = 0.42).

Cartilage Cultures

Following 24 hours of culture in the presence of cartilage explants without IL-1β, all media contained significantly higher TGFβ1 concentrations than baseline except for single-spin laboratory (P = 0.09). TGFβ1 concentrations in single-spin laboratory (1,108 ± 426 pg/mL) and kit (2,134 ± 695 pg/mL) media were not significantly different (P = 0.06). The TGFβ1 concentration in single-spin laboratory medium was significantly less than all double-spin PRP media except for double-spin 3x (2,302 ± 891 pg/mL) (P = 0.06). The TGFβ1 concentration in single-spin kit medium was not significantly different from all double-spin PRP media (P = 0.16-0.98). The TGFβ1 concentrations in double-spin 3x, 6x (2,818 ± 1,028 pg/mL), 9x (3,852 ± 2,152 pg/mL), and kit (3,552 ± 2,283 pg/mL) media were not significantly different (P = 0.17-0.75).

Meniscus Cultures

In IL-1β–free meniscus cultures, TGFβ1 concentrations increased between single-spin laboratory (806 ± 442 pg/mL) and double-spin 9x (3,427 ± 2,432 pg/mL) media, although only double-spin 6x and 9x media were higher than baseline ( Fig. 2 ). TGFβ1 concentrations in single-spin (1,071 ± 465 pg/mL) and double-spin (3,552 ± 2,283 pg/mL) kit media were not significantly different from each other (P = 0.39) and the laboratory preparations (P = 0.07-0.98).

Figure 2.

TGFβ1 concentration in media following 24 hours of culture in the absence of IL-1β. “Base” represents the concentration of TGFβ1 in laboratory PRP prior to adding to explant cultures. For each tissue, different letters indicate significant differences (P < 0.05).

ECM Synthesis

Cartilage

³ H-proline

In the absence of IL-1β, ³ H-proline incorporation in single-spin laboratory cultures was not significantly different from single-spin kit cultures (P = 0.22) ( Fig. 3 ). ³ H-proline incorporation in single-spin laboratory PRP was 31% to 64% higher than double-spin 3x, 6x, and 9x and double-spin kit cultures. In single-spin kit cultures, ³ H-proline incorporation was 31% and 20% higher than double-spin 3x and double-spin kit cultures, respectively. In double-spin kit cultures, ³ H-proline incorporation was not significantly different from double-spin 3x PRP (P = 0.25). No significant differences were found among double-spin 3x, 6x, and 9x cultures (P = 0.07-0.58). ³ H-proline incorporation decreased with the addition of IL-1β for single-spin laboratory cultures only. In IL-1β medium, ³ H-proline incorporation was not significantly different among single-spin laboratory, double-spin 3x, and single- and double-spin kit cultures (P = 0.22-0.98). ³ H-proline incorporation in double-spin 6x and 9x cultures was approximately 15% lower than single- and double-spin kit cultures.

³⁵ S-sulfate

In the absence of IL-1β, ³⁵ S-sulfate incorporation in single-spin laboratory cultures was not significantly different from single-spin kit cultures (P = 0.28) ( Fig. 3 ). ³⁵ S-sulfate incorporation in single-spin laboratory cultures was 32% to 72% higher than double-spin 3x, 6x, and 9x and double-spin kit cultures. ³⁵ S-sulfate incorporation in single-spin kit cultures was 38% to 50% higher than that in double-spin 3x, 6x, and 9x cultures. No significant differences were detected among double-spin 3x, 6x, and 9x PRP (P = 0.47-0.97). ³⁵ S-sulfate incorporation in double-spin kit cultures was not significantly different from single-spin kit and double-spin 3x, 6x, and 9x PRP cultures (P = 0.07-0.27). ³⁵ S-sulfate incorporation decreased with the addition of IL-1β for single-spin laboratory, single-spin kit, and double-spin kit cultures. In IL-1β cultures, ³⁵ S-sulfate incorporation among all media was not significantly different (P = 0.47-0.99).

Figure 3.

Radiolabel incorporation for cartilage explant cultures. The combined effects of PRP formulations and IL-1β conditions were evaluated for ³ H-proline or ³⁵ S-sulfate. For each assay, different letters indicate significant differences (P < 0.05).

Meniscus

³ H-proline

Interactions between PRP preparations and IL-1β were not significant (P = 0.78). When considering the effect of PRP independently of IL-1β, ³ H-proline incorporation in single-spin laboratory and double-spin 3x cultures was not significantly different (P = 0.46) ( Fig. 4 ). ³ H-proline incorporation in double-spin 6x and 9x cultures was not significantly different (P = 0.15), and both were significantly lower than single-spin laboratory and double-spin 3x cultures. Single-spin kit cultures were not significantly different from all other cultures (P = 0.06-0.35) except for double-spin 9x PRP. ³ H-proline incorporation in double-spin kit cultures was less than single-spin laboratory and double-spin 3x cultures only.

³⁵ S-sulfate

In the absence of IL-1β, ³⁵ S-sulfate incorporation was not significantly different among single-spin laboratory, double-spin 3x, and single-spin kit cultures (P = 0.28-0.67) ( Fig. 4B ). ³⁵ S-sulfate incorporation in double-spin 6x and 9x and double-spin kit cultures was not significantly different (P = 0.40-0.74). ³⁵ S-sulfate incorporation in double-spin 6x and 9x and double-spin kit cultures was approximately 40% less than single-spin laboratory, double-spin 3x, and single-spin kit cultures. ³⁵ S-sulfate incorporation decreased with the addition of IL-1β for single-spin laboratory, double-spin 3x, and single-spin kit PRP. In IL-1β cultures, ³⁵ S-sulfate incorporation among all media was not significantly different (P = 0.30-0.89).

Figure 4.

Radiolabel incorporation for meniscus explant cultures. For ³ H-proline, the main effect of PRP formulation was evaluated. The combined effects of PRP formulations and IL-1β conditions were evaluated for ³⁵ S-sulfate. For each assay, different letters indicate significant differences (P < 0.05).

Gene Expression of Catabolic Enzymes

Cartilage

Interactions between PRP formulation and IL-1β were not significant for MMP1 (P = 0.47), MMP13 (P = 0.12), and ADAMTS-5 (P = 0.19). When considering the effect of IL-1β, MMP1 expression increased (1.6-fold, P < 0.001) and ADAMTS-5 expression decreased (3.0-fold, P < 0.001) with IL-1β treatment (data not shown). In the absence of IL-1β, ADAMTS-4 expression in double-spin 6x cultures was 3.4- to 8.9-fold higher than single-spin laboratory, single-spin kit, and double-spin 3x cultures ( Fig. 5 ). ADAMTS-4 expression in double-spin kit cultures was not significantly different from double-spin 6x cultures (P = 0.78) and significantly higher than all other IL-1β–free cultures. The addition of IL-1β to the culture medium increased ADAMTS-4 expression for double-spin 9x cultures (11-fold) only. In IL-1β medium, ADAMTS expression was not significantly different between single-spin laboratory and double-spin 3x cultures (P = 0.97). ADAMTS-4 expression increased approximately 5-fold for double-spin 6x and 22-fold for double-spin 9x cultures. ADAMTS-4 expression in single-spin kit cultures was not significantly different from all laboratory preparations (P = 0.16-0.18) except for double-spin 9x. ADAMTS-4 expression in double-spin kit cultures was 2.1-fold higher than single-spin kit, approximately 5-fold higher than single-spin laboratory and 3x, and not significantly different from single-spin 6x (P = 0.68) cultures.

Meniscus

Expression levels of MMP1 and MMP13 were not analyzed, as many samples resulted in expression levels near the PCR detection limit. Interactions between PRP formulations and IL-1β as well as both main effects were not significant for ADAMTS-4 (P = 0.19-0.48) and ADAMTS-5 (P = 0.07-0.33) (data not shown).

Figure 5.

Normal tissue: Gene expression of ADAMTS-4 in cartilage explant cultures. Expression levels were normalized to single-spin laboratory cultures, with values less than 1 representing a relative decrease in expression. Different letters indicate significant differences (P < 0.05).

Discussion

We evaluated the influence of different PRP formulations on anabolic and catabolic activities in cartilage and meniscus explants in different PRP formulations created using laboratory processing and commercial kits. In single-spin laboratory PRP, the increase in platelet concentration over whole blood (59%) was consistent with previous reports for single-spin processing of human blood (51%). ¹⁰ Given that the vast majority of the packed cell volume (43% here) consists of red blood cells, the increase in platelet concentration in single-spin laboratory PRP over blood may be largely attributed to the replacement of the red blood cell volume with plasma and platelets during low speed centrifugation. For example, the ideal exchange of these volumes would increase the percentage of plasma and platelets in the supernatant region from 57% to 100%, for a corresponding 75% increase in platelet concentration. The platelet concentration in double-spin kit PRP (725 ± 95 thousand/µL) exceeded that of both single-spin PRPs and was comparable to a previous study (513 ± 85 thousand/µL) that produced equine PRP using the SmartPrep system (Harvest Technologies). ²⁰ The double-spin 6x and 9x PRP preparations represent higher platelet concentrations than were achieved by either commercial kit but may be obtained using other methods.

The centrifugation protocols used for laboratory processing (≤1,000g) and the single-spin kit (260g) generate forces that are at least 3 times less than a well-established protocol for human blood that does not harm platelets. ²¹ Furthermore, the pipetting used to process each preparation was less vigorous than injections through small bore needles that were found not to harm equine platelets. ²² Therefore, it is likely that little platelet activation was induced during processing of laboratory and single-spin kit PRP. Information on the centrifugation protocols used for the double-spin kit was not available from the manufacturer, although previous evaluating of PRP from this kit ²³ and the lack of significant differences in TGFβ1 concentration in uncultured PRP medium here suggest that the double-spin kit stimulates little platelet activation. The PRP medium was added to the explant cultures without performing additional steps to ensure that all growth factors were released from the platelets. We elected to test PRP in this manner, as collagen induces platelet degranulation, ²⁰ and the large collagenous articular surfaces in vivo may render an ex vivo activation step unnecessary. Also, the small amount of calcium chloride in the culture medium may have contributed to platelet activation, although typically 500-fold higher concentrations of calcium chloride (10% w/v) are used to rapidly activate platelets. The approximately 10-fold increase in TGFβ1 concentration between single-spin laboratory and double-spin 9x PRP cultures suggested that platelet degranulation occurred in explant cultures. However, TGFβ1 concentration proved to be a poor indicator of the influence of platelet growth factors on cartilage and meniscus biosynthesis given the limited number of significant differences in TGFβ1 concentration among PRP preparations. High variability in TGFβ1 concentrations in PRP has been previously reported. TGFβ1 concentration showed a poor correlation with platelet concentration in human PRP. ³ In addition, our data were consistent with TGFβ1 concentrations in unactivated equine PRP following exposure to tendon explants. ²⁰

In cartilage cultures in the absence of IL-1β, radiolabel incorporation was highest for single-spin laboratory PRP and decreased in double-spin laboratory PRP formulations. A similar trend was found for ³⁵ S-sulfate incorporation in meniscus explants cultured in the absence of IL-1β, with biosynthesis decreasing between double-spin 3x and 6x PRP. In the presence of IL-1β, no significant differences in radiolabel incorporation in cartilage and meniscal explants were detected among PRP preparations. Taken together, these data do not support the hypothesis that concentration of platelets in PRP stimulates cartilage and meniscal biosynthesis over PRP in which the platelets are minimally concentrated.

When considering the effect of IL-1β, only single-spin laboratory PRP consistently resulted in a significant decrease in biosynthesis upon the addition of IL-1β. These results were somewhat unexpected given that ³⁵ S-sulfate incorporation in cartilage explants has been reported to decrease in the presence of IL-1β,^24-26 although biosynthesis in meniscal explants has been less sensitive to IL-1β. ²⁷ While additional studies are necessary to define more detailed interactions between PRP formulation and IL-1β, it is possible that high densities of platelets produced an inflammatory response in chondrocytes, meniscus cells, or white blood cells within the PRP preparations to which 10 ng/mL of IL-1β was minimally additive. For example, ³⁵ S-sulfate incorporation in cartilage has been shown to decrease relative to IL-1β–free controls at concentrations as low as 0.01 ng/mL ²⁵ and 0.1 ng/mL.^24,26 In each of these dose titration studies, modest to no reduction in ³⁵ S-sulfate incorporation was reported between 1 and 10 ng/mL of IL-1β. Given the 10-ng/mL concentration of IL-1β used here, it appears possible that a modest proinflammatory effect of high platelet concentrations may have been sufficient to decrease biosynthesis to nearly that of the recombinant IL-1β cultures.

With the exception of ³ H-proline incorporation in meniscal explants, radiolabel incorporation in the absence of IL-1β for commercial PRP preparations coincided with biosynthesis in laboratory PRP that contained the most similar platelet concentration. Based on these data, we speculate that the single-spin procedure may be more the appropriate closed system kit of the 2 evaluated in this study for intra-articular applications. In addition, the measures of biosynthesis for double-spin 6x and 9x PRP cultures suggest that commercial kits that concentrate platelets to a greater degree than that found in this study would not provide benefit over minimally concentrated platelet preparations for intra-articular applications. These data are relevant to the choice of commercial PRP kits, as the majority of closed system kits that have been characterized for cellular content have been reported to contain at least a 2-fold concentration of platelets over whole blood.^28,29

The different PRP formulations had little effect on gene expression of catabolic enzymes associated with a proinflammatory environment, ³⁰ with significant interactions between PRP formulations and IL-1β detected for ADAMTS-4 in cartilage only. The role of ADAMTS-4 and -5 in human cartilage degradation has not been determined, ³¹ although ADAMTS-4 and -5 have been associated with diseased equine cartilage ³² ; therefore, the observed differences in ADAMTS-4 expression here may be suggestive of an important catabolic process in equine cartilage. Based on the laboratory preparations resulting in an increase in ADAMTS-4 gene expression between double-spin 3x and 6x PRP, these data further suggest that high platelet concentrations in PRP may actually produce a proinflammatory environment for cartilage.

In single-spin kit PRP cultures, ADAMTS-4 expression was not significantly different from single-spin laboratory PRP. Conversely, double-spin kit PRP cultures resulted in ADAMTS-4 expression that was higher than double-spin 3x PRP in which the platelet concentration was most similar. A major difference in cellular composition of PRP preparations that may account for this result is the high concentration of white blood cells in double-spin kit PRP (14.8 ± 3.0 thousand/µL) that exceeds the normal white blood count that we have historically recorded for young adult horse blood in our laboratory (<5 thousand/µL). It has been suggested that the leukocyte population in PRP enriched with white blood cells is proinflammatory, ³³ a theory that is supported by a report of increased MMP expression in tendon explants cultured in white blood cell–rich PRP. ²⁰ While the influence of white blood cells in PRP for clinical applications remains controversial, ³⁴ and additional studies are necessary to better understand the influence of leukocytes in PRP on catabolism, our data suggest that white blood cells in PRP may be proinflammatory for cartilage.

This study evaluated PRP formulations across a range of platelet concentrations as an indicator of whether differences exist among the various PRP kits that are commercially available. A limitation of this study was the lack of a synovial fluid control group, which would have helped support or refute the hypothesis that PRP enhances biosynthesis over the natural joint environment. In addition, when we project the results of in vitro studies to clinical applications, certain limitations of laboratory models must be considered. While the dilution of PRP by 50% was intended to mimic the dilution of an intra-articular injection in synovial fluid, tissue culture does not allow for clearance of the growth factors ³⁵ as may occur in joints. Therefore, the laboratory model may overestimate the concentration of growth factors that is present in vivo. Based on our growth factor analysis, equine platelets contain less TGFβ1 than has been reported for human preparations, although the extent to which other platelet-derived factors may differ among species is unknown. Most importantly, the acute response of cartilage and meniscal explants does not necessarily recapitulate the response that occurs over weeks of clinical healing. While in vivo testing will be necessary to address these limitations, the present study rejects the hypothesis that the concentration of platelets in PRP stimulates ECM synthesis in cartilage and the meniscus; furthermore, these findings suggest that high platelet concentrations for intra-articular injection should be considered with caution.