Since sperm vitrification in straws was first described in human by Isachenko et al. (2005 ), it has been investigated in different animals including ram ( Jiménez-Rabadán et al. , 2015 ), goat ( Daramola et al. , 2016 ), fish ( Zilli et al. , 2018 ) and horse ( Consuegra et al. , 2018 ; Restrepo et al. , 2019 ). This technique has also been developed in donkey sperm in absence of permeable cryoprotectants, showing promising results regarding in vitro parameters ( Diaz-Jimenez et al. , 2017 , 2018b ). In this sense, vitrification in straws would be an interesting alternative to conventional slow freezing in this species, whose results after ar­tificial insemination in jennies using frozen sperm are very poor, with pregnancy rates ranging from 0 to 36% ( Rota et al. , 2012 ; Oliveira et al. , 2016 ). However, the relatively small volume employed (100 µL) seems to be a major disadvantage in comparison to conventional freezing ( Slabbert et al. , 2015 ). To solve this limitation, vitrification of large volumes of sperm (up to 0.5 mL) has been developed in human beings, showing similar ( Slabbert et al. , 2015 ) and higher ( Isachenko et al. , 2011 ) sperm motility parameters after warming when compared to conventional freezing. Nevertheless, to our knowledge, there is little information about its use in animals, and if so, no motile sperm was recovered after warming ( Restrepo et al. , 2019 ).

Temperature and time during thawing have shown to affect sperm motility, plasma membrane and acrosome integrity after conventional freezing ( Athurupana et al. , 2015 ); and they seem to be also key factors for sperm vitrification success ( Mansilla et al. , 2016 ). A few studies comparing warming temperatures after vitrification could be found in the literature: Mansilla et al. (2016 ) reported optimum results in vitrification of human sperm in 0.25 mL straws after warming at 42ºC in comparison to lower temperatures; however, heterogeneous results were obtained after sperm vitrification following the spheres method. In this sense, high warming temperatures (up to 60ºC) seemed to be more adequate for mouflon ( Pradiee et al. , 2016 ) and ibex ( Pradiee et al. , 2015 ) than for dog ( Caturla-Sánchez et al. , 2018 ) sperm, in which slower warming rates (37ºC) showed better sperm quality after vitrification. To the best of our knowledge, there are no previous studies comparing different temperatures and times for warming after vitrification in donkey sperm. Therefore, the aim of this study was to vitrify donkey sperm in 0.5 mL straws using different temperatures and rates for warming in comparison to conventional freezing regarding sperm motility parameters, plasma membrane and acrosome integrity.

Sperm parameters before cryopreservation had the following average values: gel-free volume = 61.1 ± 25.3 mL, sperm concentration = (354.5 ± 101.2) × 10 ⁶ sperm/mL, TM = 89.5 ± 6.4%, PM = 61.0 ± 15.7% and PMI = 61.8 ± 12.7%. Conventional freezing showed significantly ( p <0.001) higher results than vitrification for TM = 42.7 ± 19.6%, PM = 30.3 ± 16.7%, PMI = 49.1 ± 10.4% and AIS = 39.6 ± 14.5% despite the warming procedure employed (Fig. 1). No differences ( p >0.05) were observed among warming temperatures after vitrification for TM (W1: 0.2 ± 0.4%, W2: 2.8 ± 4.3% and W3: 2.4 ± 2.1%) and PM (W1: 0.2 ± 0.4%, W2: 1.2 ± 1.8% and W3: 2.1 ± 1.9%). However, PMI was significantly higher in W3 (20.0 ± 11.0%) than W1 (5.5 ± 3.6%) and W2 (9.3 ± 8.4%) ( p <0.05). The percentage of AIS resulted also in higher values in W3 (29.8 ± 4.5) than W1 (14.5 ± 3.7) ( p <0.05).

Parameters regarding velocity features showed no differences ( p >0.05) between W3 and C for VCL VSL and VAP. Protocol W1 resulted in the lowest values for LIN, STR, WOB, ALH and BCF. Mean values of kinematic sperm parameters assessed are represented in Table 1.

Vitrification of sperm in large volumes has been mostly researched in human showing similar ( Slabbert et al. , 2015 ) or higher ( Isachenko et al. , 2011 ) sperm quality after warming when compared to conventional freezing. However, in the present study conventional freezing conserved motility, plasma membrane and acrosome integrity better than vitrification in 0.5 mL straws in donkey sperm. A possible explanation for this difference is the higher cryostability of the human sperm due to its structural composition ( Isachenko et al. , 2003 ). Cryostability of sperm cells has been negatively correlated with the size, probably due to higher water content and lower internal compaction of the bigger ones ( Katkov et al. , 2012 ). Human sperm head size is smaller in comparison to other mammalian species, including equine ( Garner, 2006 ; Jiménez-Rabadán et al. , 2015 ; Arraztoa et al. , 2017 ). This characteristic probably makes human sperm less vulnerable to cryopreservation damage ( Arraztoa et al. , 2017 ), which allows successful vitrification in hi­gher volumes. Additionally, differences between these species exist also in plasma membrane composition of the sperm (cholesterol and phospholipid types) and they are also determinants of sperm cryostability. Besides, sucrose concentration employed in the present study was selected from preliminary tests in donkey sperm vitrification in spheres (30 µL), in which 0.1 M sucrose conserved sperm parameters better than the extender commonly used for human sperm vitrification: 0.25 M sucrose prepared in distilled water plus serum albumin ( Isachenko et al. , 2008 ; Sanchez et al. , 2012 a , b ). In this sense, the extender requirements for guaranteeing vitrification of donkey sperm packed in 0.5 mL straws may be different in comparison to spheres method. Therefore, further studies testing higher non-permeable cryoprotectant concentration; other sugars, such as trehalose ( El-Badry et al. , 2017 ; Schulz et al. , 2017 ); as well as other substances supplementation, such as cyclodextrins ( Madison et al. , 2013 ; Partyka et al. , 2016 ; Lone, 2018 ) would be interesting to improve results after sperm vitrification in 0.5 mL straws in this species.

In addition, human sperm vitrification has been traditionally performed after sperm selection of motile and viable cells; whereas sperm selection techniques prior to freezing in stallions has shown no improvements on sperm motility after thawing ( Hidalgo et al. , 2017 ). In this regard, it is also important to point out that vitrification in 0.5 mL straws has obtained despair results regarding species, even among studies in human sperm, as previously described by Katkov et al. (2012 ). Conventional freezing has shown better results than sperm vitrification in other domestic animals as dog ( Caturla-Sánchez et al. , 2018 ), and ram ( Pradiee et al. , 2016 ); but the spheres method was used in both.

In order to maximize the possibilities of sperm vitrification success in large volumes in donkeys, different warming protocols were compared. Unex­pectedly, no differences in motility percentages bet­ween warming rates were found; although higher temperatures during shorter periods of time (W3, 70ºC/8s + 37ºC/52s), showed significantly higher membrane integrity and velocity features than lower temperatures during longer periods of warming (W1 and W2). Moreover, W3 warming protocol outcomes similar sperm velocity parameters (VSL, VCL and VAP) after vitrification than those obtained after conventional freezing. In this regard, hyperactivation of sperm cells might be suspected; however, it has been generally defined as a change in the pattern of sperm motility ( Hinrichs & Loux, 2012 ), which has been set in stallion sperm as (i) VCL>180 µm/s and mean ALH>12 µm ( Rathi et al. , 2001 ) or (ii) VCL between 225 and 350 µm/s and ALH between 8 and 13 µm ( Hinrichs & Loux, 2012 ). In this sense, none of the kinematic sperm parameters assessed in this study had those characteristics, so sperm hyperactivation after vitrification could be dismissed. On the other hand, sperm velocity parameters ascertained by CASA assessments have been previously correlated with in vivo fertility in mammals ( Gomendio & Roldan, 2008 ). Unfortunately, fertility has not been assessed in this study.

Additionally, the better sperm quality found in the present study using high warming rates (W3), is in agreement to those results previously reported by Pradiee et al. (2016 ), in which rapid warming prevented damage after vitrification in spheres in mouflon sperm (60ºC better than 37ºC). Similarly, Athurupana et al. (2015) found higher motility and membrane integrity sperm values using 70ºC/8s + 39ºC/52s in comparison to 39ºC/60s in conventional freezing of boar sperm. On the contrary, a recent work comparing warming procedures after dog sperm vitrification in spheres ( Caturla-Sánchez et al. , 2018 ), showed better sperm quality after warming at 37ºC rather than 65ºC. In the present study, almost no motile sperm were recovered using the conventional thawing protocol for 0.5 mL French straws (W1, 37ºC/30s), which may be explained because the warming rate for sperm vitrification in large volumes must be raised in comparison to conventional freezing. In this sense, W2 (43ºC/20s) was expected to outcome good sperm quality after vitrification because similar warming protocols were successfully employed in donkey ( Diaz-Jimenez et al. , 2017 , 2018b ) and human ( Mansilla et al. , 2016 ) sperm vitrification using 0.25 mL straws. Nevertheless, volume and package were modified in this report in comparison to the previous ones, so the vitrification procedure was completely different. The ratio between volume and surface occupied by the semen samples in 0.5 mL straws was different in comparison to 0.25 mL straws, reducing thereby the cooling rates achieved in this study. According to this, the ultra-rapid cooling rates required for successful ‘kinetic vitrification' of equine sperm may not occur ( Hidalgo et al. , 2018 ), which could explain the poor sperm quality obtained after warming.

Sperm motility after vitrification yielded substan­ti­ally lower percentages in comparison to conventional freezing, being therefore deficient for its use in routine artificial insemination practices. However, percentage of sperm viability using the W3 warming protocol was higher than other warming rates tested, reaching values around 20%. In this regard, selection of sperm cells after warming might be performed ( Ortiz et al. , 2015 ) and possibly be used for advanced reproduction techniques such as "intracytoplasmic sperm injection" (ICSI) ( Gonzalez-Castro & Carnevale, 2019 ).

In conclusion, high warming rates seem to be better for donkey sperm vitrification in large volumes; but this methodology is still not an alternative to conventional sperm freezing.

The authors are extremely grateful to Centro Rural Malpica (Palma del Río, Cordoba, Spain) for providing the animals.