A field investigation of intra-cervical insemination with reduced sperm numbers in gilts

A novel insemination catheter with a smaller polyurethane tip for deeper insertion into the cervix of gilts was compared with the conventional catheter. The novel catheter could be inserted 31.4 mm deeper than the conventional catheter into the gilt cervix, but the difference diminished with parity until the sixth parity when there was no difference in penetration depth between the catheters. In Experiment 1, cyclic gilts were inseminated upon display of oestrus (back pressure test) in the presence of a boar (0 h) and 24 h later. The control group (n = 300) were inseminated with 2 x 10(9) total spermatozoa and the treatment group (n = 300) with 1 x 10(9) total spermatozoa per inseminate, in both cases utilising the novel insemination catheter. No significant differences were observed for farrowing rate and litter size, the values of which were those expected for natural mating. In Experiment 2, 66 cyclic gilts were subjected to the same heat detection and service regime as for Experiment 1 but were served with <1 x 10(9) total sperm cells per inseminate using the new device. Conception rates and embryo counts were recorded. Conception rate declined with <500 x 10(6) spermatozoa, and number of embryos (a reflection of potential litter size) was significantly reduced. Use of the new catheter for gilts with 1 x 10(9) total sperm cells per inseminate will achieve commercially acceptable fertility and fecundity levels, and offer substantial commercial benefits with more rapid genetic gains.     

Effect of insemination of estrus-induced prepuberal gilts on ensuing reproductive performance and body weight

Prepuberal gilts reared and managed to 85-90 kg live weight in a common system were allocated at random to one of three first-mating treatments in an experiment conducted over a period of more than 5 years. In two of the treatments, gilts received a single i.m. injection of 400 IU equine chorionic gonadotropin (eCG) and 200 IU human chorionic gonadotropin (hCG) (PG600; Intervet) and were either inseminated 4 and 5 days later on a fixed-time basis regardless of oestrus (treatment A), or at the second oestrus following treatment (treatment B). The third group of gilts remained untreated and was inseminated on the first spontaneous oestrus (treatment C). Thereafter, all gilts were managed in the same way and those observed in oestrus were re-inseminated. Significantly more gilts returned to oestrus after the first service in treatment A (35%) than in treatment B and C (12 and 17%, respectively; P<0.01). Gilts farrowed to the first or repeat inseminations at a significantly younger age (P<0.01) in treatment A (304 days) than treatment B (324 days) and C (320 days). The age difference at farrowing remained in surviving gilts at the end of their third parity. The first farrowing performance of the gilts was significantly affected by treatment in terms of litter size at birth (A 7.0, B 8.4 and C 8.3 live piglets per gilt; P<0.01), litter size at weaning (A 6.2, B 7.2 and C 7.2 live piglets per gilt, P<0.05), and piglet birth weight (A 1.4, B 1.3 and C 1.3 kg; P<0.05) but piglet survival rate and weaning weight were not affected by treatment. The live weights of the gilts were significantly different between the treatments at first insemination (A 95.7, B 106.5 and C 109.2 kg; P<0.01) but not when the first litter was weaned (A 133.6, B 135.1 and C 136.6; P>0.05). After the first farrowing there were no differences between the treatments in terms of the survival rate, productive or reproductive performance of the gilts/sows and their offspring. Without conducting a detailed cost-benefit-calculation it was deduced that, from an economical point of view, differences between treatment A and treatments B and C are negligible because the savings associated with farrowing at a younger age on this treatment just about compensated for any additional costs associated with the treatment and the lower number of piglets born at the first farrowing.     

Effect of side of insemination on transuterine transport of spermatozoa in superovulated dairy cattle

Eleven superovulating Friesian lactating cows were inseminated deep into one uterine horn with one unit of frozen semen, containing 2.5 million total spermatozoa, with more than 40% postthaw progressive motility and with 14% morphologically abnormal spermatozoa. Semen was deposited into the right or left uterine horns of alternate cows. There was no difference in the proportions of fertilized recovered ova from ipsilateral horns between right and left inseminations (P>0.05). The fertilization rate in the contralateral horns was higher (P<0.01) for right uterine horn insemination (50%) than for left uterine horn insemination (15.6%).     

Inhibition of pregnancy with indomethacin in mature gilts and prepuberal gilts induced to ovulate

Twenty prepuberal (P) gilts, 56.5 +/- 1.1 kg body weight, were induced to ovulate with 1000 IU of pregnant mare's serum gonadotropin followed 72 h later by 500 IU of human chorionic gonadotropin (hCG), and bred by artificial insemination (AI) with 50 ml fresh pooled boar semen the day after hCG treatment (Day 0). Eighteen mature (M) gilts, 120.6 +/- 1.7 kg body weight, were bred by AI each day of estrus using pooled semen from the same boars (onset of estrus = Day 0). One-half of each group was fed the prostaglandin (PG) synthesis inhibitor indomethacin (IND), at 10 mg/kg body weight, or control (C) feed twice daily on Days 10 to 25. Blood samples taken by venipuncture on Days 10, 15, 20 and 25 were quantitated for progesterone (P4) and 13,14-dihydro-15-keto-PGF2 alpha (PGFM) by radioimmunoassay. Ovaries were examined on Day 26. All M-C gilts were pregnant, whereas 3 of 9 M-IND gilts (P less than 0.05) and none of the P gilts (P less than 0.01) were pregnant. Three of the 6 nonpregnant M-IND gilts displayed estrus on Day 21. The 3 remaining M-IND gilts had maintained corpora lutea (CL) on Day 26. Only corpora albicantia were present in P gilts on Day 26. Serum P4 concentrations for M-C gilts, nonpregnant M-IND gilts with maintained CL, and pregnant M-IND gilts were not different. Serum P4 for all nonpregnant gilts in which CL had regressed by Day 25 decreased to less than 5 ng/ml on Day 20.(ABSTRACT TRUNCATED AT 250 WORDS)     

Embryo production from superovulated Holstein-Friesian dairy heifers and cows after insemination with frozen-thawed sex-sorted X spermatozoa or unsorted semen

The aim of this study was to evaluate embryo production in superovulated Holstein-Friesian dairy heifers and cows inseminated with either X-sorted spermatozoa (2 million/dose) or unsorted semen (15 million/dose). Experiment 1 at the research farm involved eight heifers, six cows and semen of one Holstein bull. All transferable embryos were diagnosed for sex. Experiment 2 included embryo collections on commercial dairy farms: X-sorted spermatozoa from three Holstein bulls were used for 59 collections on 28 farms and unsorted semen from 32 Holstein bulls were used for 179 collections on 79 farms. Superovulations were induced by eight declining doses of FSH (total of 12 ml for heifers and 19 ml for cows) starting on days 8-12 of the estrus cycle. Inseminations began 12h after the onset of estrus and were performed two to four times at 9-15 h intervals. Low-dose X-sorted inseminates were deposited into uterine horns and unsorted semen was placed into the uterine body. In Experiment 1, on average 70.3 and 75.0% of embryos recovered from heifers, and 48.4 and 100% of embryos recovered from cows were of transferable quality in X-sorted and unsorted groups, respectively. The proportion of transferable female embryos produced approximately doubled when insemination was with X-sorted spermatozoa compared to insemination with unsorted semen (heifers 96.4% versus 41.1%; cows 81.1% versus 39.8%). In Experiment 2, estimated 53.9 and 65.5% of embryos recovered from heifers, and 21.1 and 64.5% of embryos recovered from cows were of transferable quality in X-sorted and unsorted groups, respectively. Proportions of unfertilized oocytes were 21.1 and 10.6% for heifers and 56.0 and 14.4% for cows in X-sorted and unsorted groups, respectively. Consequently, cows inseminated with X-sorted spermatozoa produced significantly smaller proportions of transferable embryos (p<0.005) and significantly larger proportions of unfertilized oocytes (p<0.001) than those inseminated with unsorted semen. Proportions of quality 1 or degenerated embryos were similar for the two treatments in both heifers and cows. Within treatments, bulls did not significantly affect the proportions of transferable, unfertilized or degenerated oocytes/embryos. It was concluded that using low-dose X-sorted spermatozoa rather than normal-dose unsorted semen for the insemination of superovulated embryo donors can improve the proportion of transferable female embryos produced but this potential may not be achieved in commercial practice, particularly in cows, because of reduced fertilization rates when using low doses of X-sorted spermatozoa.     

Insemination of mares with low numbers of either unsexed or sexed spermatozoa

Two experiments were conducted to determine pregnancy rates in mares inseminated 1) with 5, 25 and 500 x 10(6) progressively motile spermatozoa (pms), or 2) with 25 x 10(6) sex-sorted cells. In Experiment 1, mares were assigned to 1 of 3 treatments: Group 1 (n=20) was inseminated into the uterine body with 500 x 10(6) pms. Group 2 (n=21) and Group 3 (n=20) were inseminated into the tip of the uterine horn ipsilateral to the preovulatory follicle with 25 and 5 x 10(6) pms, respectively. Mares in all 3 groups were inseminated either 40 (n=32) or 34 h (n=29) after GnRH administration. More mares became pregnant when inseminated with 500 x 10(6) (18/20 = 90%) than with 25 x 10(6) pms (12/21 = 57%; P<0.05), but pregnancy rates were similar for mares inseminated with 25 x 10(6) vs 5 x 10(6) pms (7/20 = 35%) (P>0.1). In Experiment 2, mares were assigned to 1 of 2 treatments: Group A (n=11) was inseminated with 25 x 10(6) spermatozoa sorted into X and Y chromosome-bearing populations in a skimmilk extender. Group B (n=10) mares were inseminated similarly except that spermatozoa were sorted into the skimmilk extender + 4% egg yolk. Inseminations were performed 34 h after GnRH administration. Freshly collected semen was incubated in 224 microM Hoechst 33342 at 400 x 10(6) sperm/mL in HBGM-3 for 1 hr at 35 degrees C and then diluted to 100 x 10(6) sperm/mL for sorting. Sperm were sorted by sex using flow cytometer/cell sorters. Spermatozoa were collected at approximately 900 cells/sec into either the extender alone (Group A) or extender + 4% egg yolk (Group B), centrifuged and suspended to 25 x 10 sperm/mL and immediately inseminated. Pregnancy rates were similar (P>0.1) between the sperm treatments (extender alone = 13/10, 30% vs 4% EY + extender = 5/10, 50%). Based on ultrasonography, fetal sex at 60 to 70 d correlated perfectly with the sex of the sperm inseminated, demonstrating that foals of predetermined sex can be obtained following nonsurgical insemination with sexed spermatozoa.     

Hysteroscopic insemination of small numbers of spermatozoa at the uterotubal junction of preovulatory mares

Mares were inseminated with motile spermatozoa suspended in 30-150 microliters Tyrode's medium directly onto the uterotubal papilla at the anterior tip of the uterine horn, ipsilateral to the ovary containing a dominant preovulatory follicle of > or = 35 mm in diameter, by means of a fine gamete intrafallopian transfer (GIFT) catheter passed through the working channel of a strobed light videoendoscope. Insemination of 10, 8, 25, 14, 11 and 10 mares with, respectively, 10.0, 5.0, 1.0, 0.5, 0.1 or 0.001 x 10(6) motile spermatozoa resulted in conception rates of, respectively, 60, 75, 64, 29, 22 and 10%. Deposition of 1.0 x 10(6) motile spermatozoa onto the uterotubal papilla began to approach the limit of successful fertilization. These doses are far lower than the 3-15 x 10(9) spermatozoa normally ejaculated by fertile stallions during mating, and the accepted minimum dose of 500 x 10(6) spermatozoa used for conventional uterine body insemination in mares. The simplicity of the technique offers a practical means of exploiting new breeding technologies that require very small numbers of spermatozoa in horse breeding.     

Prostaglandin F concentrations in utero-ovarian vein plasma of prepuberal and mature gilts

Prostaglandin F (PGF) and progestins in utero-ovarian vein (UOV) plasma during the late luteal phase of the estrous cycle in unbred mature gilts and following induced ovulation in unbred prepuberal gilts were determined. Prepuberal gilts (120 to 130 days of age) were induced to ovulate with Pregnant Mare Serum Gonadotropin and Human Chorionic Gonadotropin (HCG). The day following HCG was designated as Day 0. Mature gilts which had displayed two or more estrous cycles of 18 to 22 days (onset of estrus = Day 0) were used. Polyvinyl catheters were inserted into the UOV of all gilts and blood was collected at 15 min intervals from 0800 to 1045 hr on Days 10 through 20 or Days 12 through 18. Plasma PGF concentrations in the mature gilts were elevated on Days 13, 14, 15, 16 and 17, whereas, plasma PGF concentrations in the prepuberal gilts were elevated only on Days 15, 16 and 17 resulting in a reproductive age (mature vs prepuberal) by day interaction (P<.01). In addition, the PGF concentrations on Days 13 through 17 were consistently greater in the mature gilts than in the prepuberal gilts as was the overall mean PGF concentration (1.95 vs .83 ng/ml). The average peak PGF concentration throughout the sampling period (4.6 vs 2.5 ng/ml; P<.01) and the average peak PGF concentration prior to luteal regression (3.8 vs 1.1 ng/ml; P<.05) were also greater in the mature than in the prepuberal gilts. Based on these results, we suggest that luteal regression in the bred prepuberal gilt following induced ovulation may not be due to an excessive production of the uterine luteolysin, but rather that the induced corpora lutea (CL) of the prepuberal gilt may be more sensitive to the uterine luteolysin than the spontaneously formed CL of the mature gilt.     

Intrauterine insemination enhances fertility of frozen semen in superovulated ewes

Superovulated ewes were inseminated with fresh or frozen semen in a factorial experiment which compared two techniques of artificial insemination; i.e. conventional cervical deposition and intrauterine deposition at laparoscopy. Similar fertilization rates resulted from insemination with fresh semen at cervical (81% of ova from 11/11 ewes) and intrauterine (83% of ova from 10/12 ewes) sites. These results approached those observed in a naturally-mated group (95% of ova from 5/5 ewes). In ewes inseminated with frozen semen, fertilization rate was markedly reduced (P less than 0.05) after cervical insemination (11% of ova from 3/11 ewes) and partly restored (P less than 0.05) after intrauterine insemination (50% of ova from 8/11 ewes).     

Characterization of uterine leukocyte infiltration in gilts after artificial insemination

The objective of this study was to characterize the uterine leukocyte influx after artificial insemination (AI). After detection of oestrus with a boar at intervals of 1.5 h, seventy-two gilts were randomly assigned to a 2 x 3 x 4 factorial arrangement. AI was performed with 100 ml extended semen containing 5 x 10(9) spermatozoa (semen; n = 36) or 100 ml VSP semen extender (extender; n = 36) at one of three times after detection of oestrus: 12, 24 or 36 h (n = 24/time). The uterus was lavaged at 6, 12, 18 or 24 h (n = 18/time) after AI to determine the total number of uterine leukocytes. In addition, uterine lavage was performed on nine untreated gilts immediately after the detection of oestrus to establish a baseline number of leukocytes. The leukocyte response in all samples consisted predominately (92-99%) of polymorphonuclear neutrophilic granulocytes (PMNs). The mean number of PMNs recovered from the uteri of gilts treated with semen was greater than in gilts treated with extender and in untreated gilts (P < 0.01). The greatest number of PMNs in semen-treated gilts was found 12 h after AI (P < 0.01), and this number was sustained for 24 h. In contrast, the number of uterine PMNs recovered from extender-treated gilts reached a peak at 6 h and had declined by 12 h after AI (P < 0.05). It was concluded that an extensive influx of PMNs into the uterus is a normal sequence to AI. The consequences and importance of semen-induced uterine leukocytosis needs further investigation.     

Sperm survival and heterogeneity are correlated with fertility after intrauterine insemination in superovulated ewes

Efficient animal production involves accurate estimations of fertilizing ability. One key factor is the plasma membrane of the sperm cell, which is actively involved in the cascade of events before oocyte fusion. Many methods are used to analyze the characteristics of this membrane, including partition in aqueous two-phase systems which is an efficient method to analyze sperm surface changes accounting for loss of viability and different functional states. Centrifugal countercurrent distribution (CCCD) analysis can also be used in an aqueous two-phase system to determine the relationship between sperm parameters and in vivo fertility in ewes. In a previous work, we found a significant correlation between two post-CCCD parameters (heterogeneity and recovered viability) and field fertility when the same sample was used after cervical AI. The present study was intended to find out whether the control of several external factors that affect reproductive efficiency is able to increase the correlation coefficient between post-CCCD parameters and fertility. Thus, 90 Rasa aragonesa ewes were controlled on the same farm and received intrauterine inseminations using the same technical equipment. The fertilizing ability of the raw semen and sperm samples selected by a dextran/swim-up process was compared using a low number of spermatozoa per insemination (7 x 10(7)) to enhance possible fertility differences. A new post-CCCD parameter was considered; the loss of viability (LV) occurred during the CCCD process. This variable denotes the sperm surviving ability and corresponds to the difference between the total number of viable cells loaded and recovered after the CCCD run. The mean fertility of eight sperm control samples was 60% (range: 25-76%), and there was no significant correlation between standard parameters and in vivo fertility. LV ranged from 2 to 69% (average 27%) and was negatively correlated with fertility (r = -0.914, P < 0.01). Ejaculate heterogeneity (H) ranged from 20 to 47% and was positively, but not significantly, correlated with fertility (r = 0.391). A predictive equation for fertility was deduced by multiple analysis with a very high correlation coefficient (r = 0.967), and level of significance (P < 0.005): predictive fertility PF = 52.546 - 0.594 LV + 0.665 H. The mean fertility of 13 swim-up selected samples was 63% (range: 25-86%). Again, only parameters derived from the CCCD analysis were highly correlated with fertility, especially LV and H (P < 0.05).     

Premature elevation of progesterone shortens duration of ovulation in PMSG/hCG-treated prepuberal gilts

A surge of LH during the follicular phase triggers multiple pathways, including progesterone and prostaglandin synthesis before culminating in ovulation. Progesterone has been shown to be involved in the ovulatory process in many species. In prepuberal gilts treated with PMSG/hCG the follicular progesterone level has been shown to increase sharply before ovulation. This study was conducted to investigate whether premature elevation of progesterone can accelerate the ovulatory process in Large White PMSG/hCG-treated prepuberal gilts. Fifty-four Large White gilts were treated with 1000 IU, i.m. PMSG to stimulate follicular growth, followed 72 h later by 500 IU, i.m. hCG to induce ovulation. Gilts in the treatment group (n = 27) were given progesterone intermuscularly at 24 and 36 h after hCG. Ovaries were exteriorized to observe ovulation points during laparotomy under general anesthesia at 38 to 50 h after hCG. Ovulation in both groups commenced by 40.05 h after hCG and was completed by 47.71 h in the control group and by 42.87 h after hCG in the treated group. Progesterone shortened (P < 0.01) ovulation time by 4.84 h and the time required (P < 0.01) for the median proportion of follicles to ovulate (40.7 vs 43.5 h after hCG). Progesterone also increased (P < 0.01) the plasma progesterone concentration without altering follicular progesterone concentration.     

Pregnancy rates in mares following hysteroscopic or transrectally-guided insemination with low sperm numbers at the utero-tubal papilla

This study was conducted to evaluate two methods for insemination of a low number of sperm in the tip of the uterine horn, and to determine whether prebreeding intrauterine treatment with prostaglandin E(2) would improve pregnancy rates. Estrus was synchronized in 36 fertile Quarter Horse and Thoroughbred broodmares. When a dominant follicle >or=33 mm diameter was present, mares were treated with 2500 units hCG intravenously and were assigned to one of four treatment groups for insemination with five million total sperm in 200 microl extender the next day as follows: (1) Group PGE-HYS (n=9): 0.25mg PGE(2) in 1 ml 0.9% NaCl solution infused into the tip of the uterine horn ipsilateral to the dominant follicle 2h prior to hysteroscopic-guided inseminate placement onto the oviductal papilla; (2) Group SAL-HYS (n=9): 1 ml 0.9% NaCl solution infused into the tip of the uterine horn ipsilateral to the dominant follicle 2h prior to hysteroscopic-guided inseminate placement onto the oviductal papilla; (3) Group PGE-PIP (n=9): 0.25mg PGE(2) in 1 ml 0.9% NaCl solution infused into the tip of the uterine horn ipsilateral to the dominant follicle 2h prior to transrectally-guided pipette placement of the inseminate into the tip of the uterine horn; and (4) Group SAL-PIP (n=9): 1 ml 0.9% NaCl solution infused into the tip of the uterine horn ipsilateral to the dominant follicle 2h prior to transrectally-guided pipette placement of inseminate into the tip of the uterine horn. Mares in estrus were evaluated daily by transrectal ultrasonography to monitor follicular status and confirm ovulation. If mares had not ovulated within 2 days of insemination, the assigned treatment was repeated. Pregnancy status was evaluated by transrectal ultrasonography 12-14 days postovulation, and pregnancy rates were compared.No interaction between prebreeding treatment (SAL:PGE) and insemination protocol (HYS:PIP) on pregnancy rates occurred (P>0.10). Pregnancy rates did not differ between mares inseminated by HYS (12/18; 67%) or PIP (10/18; 56%) (P>0.10). Pregnancy rates did not differ between mares treated prior to breeding with PGE (11/18; 61%) or SAL (11/18; 61%) (P=1.00). In summary, satisfactory pregnancy rates were obtained when a low number of sperm were either placed directly onto the oviductal papilla using hysteroscopy or placed in the tip of the uterine horn using a transrectally-guided uterine pipette. Infusion of 0.25mg PGE(2) in the tip of the uterine horn 2h prior to insemination did not improve pregnancy rates.     

Comparisons between three different extenders for canine intrauterine insemination with frozen-thawed spermatozoa

Ejaculates from 7 dogs were obtained on the same day and were pooled. This pooled semen was separated into 3 equal fractions and processed simultaneously, the only difference being in the extender used for freezing. The extenders were laiciphos (containing laiciphos, egg yolk, distilled water and glycerol- Group 1); Tes/Tris (containing Tes/Tris, egg yolk, distilled water and glycerol- Group 2); and biociphos (containing biociphos with glycerol in it, egg yolk and distilled water- Group 3). Spermatozoa were conditioned in 0.5ml French straws and presented normal characteristics before freezing and after thawing. The sperm concentration of the pooled was 683 x 10(6) sperm/ml; sperm motility was above 95%, the percentage of live spermatozoa was above 95% and was of good quality and mobility. Characteristics of the spermatozoa after thawing were the same for spermatozoa frozen with laiciphos and Tes/Tris. Mean sperm concentration was 201.5 +/- 4.95 x 10(6) sperm/ml, sperm motility was 65%, the percentage of live spermatozoa was 80% and the quality of motility.was good. Spermtozoa frozen with biociphos had the following post-thaw characteristics: sperm concentration was 201 x 10(6) sperm/ml, sperm motility was 50%, the percentage of live spermatozoa was 78% and the quality of mobility was medium. Abnormalities were less than 15% for all spermatozoa after thawing. Intrauterine artificial inseminations were performed by laparoscopic intrauterine insemination twice at Days 3 and 5 after the estimated LH peak in 15 normally cyclic Beagle bitches (5 per group) presenting normal hormonal profiles. There were no differences between groups. The females were inseminated with 1.0 ml of spermoatozoa (concentration of 200 x 10(6) sperm/ml) diluted with 1.0 ml of extender. A 60% pregnancy rate was obtained in bitches inseminated with frozen-thawed spermatozoa extended with laiciphos or Tes/Tris and 100% in bitches inseminated with spermatozoa extended with biociphos. Females inseminated with laiciphos, Tes/Tris and biociphos had a mean litter size of 5 +/- 2.6, 3 +/- 1 and 3.4 +/- 1.3 pups, respectively. This study demonstrated that post-thaw assessment of sperm characteristics is not the best technique for evaluating sperm fertility after freezing or for assessing different semen extenders.     

Impact of Percoll on bovine spermatozoa used for in vitro insemination

Percoll treatment of bovine frozen/thawed spermatozoa was identified as the cause of low cleavage and blastocyst rates in our in vitro embryo production system. Percoll treatment of spermatozoa yielded very high, stable and repeatable results for many months of cleavage and blastocyst rates of 83 +/- 5% (+/-SD) and 33 +/- 10%, respectively. This was followed by a period with significantly lower cleavage and blastocyst rates of 60 +/- 9% and 14 +/- 5%, respectively. In the last period, the results became even lower, with cleavage and blastocyst rates of 32 +/- 5% and 5 +/- 3%, respectively. These results were significantly different from those of the 2 previous periods. We were able to correlate precisely these changes in outcomes to the introduction of new Percoll batches. When compared with washed spermatazoa and under otherwise identical circumstances, the cleavage and blastocyst rates from spermatozoa treated with Percoll increased significantly from 33 to 73% and from 4 to 26%, respectively. We suggest that this adverse effect of Percoll is not due to Percoll particles per se, but may be ascribed to the effect of unbound polyvinylpyrrolidone (PVP) in the Percoll.