| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Department of Anatomy and Cellular Biology and
b Department of Obstetrics and Gynecology, Tufts UniversitySchool of Medicine and New England Medical Center, Boston, Massachusetts 02111
c Center for Research on Reproduction&Women's Health,
d Department of Biology, University of Pennsylvania, Philadelphia, Pennsylvania 19104
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Numerous investigators have demonstrated that ovulated eggs display a time-dependent increase in their susceptibility to be parthenogenetically activated [8–12], and we previously demonstrated that this susceptibility is correlated with a time-dependent increase in the transition of the ovulated MII-arrested egg to an interphase-like state [13]. This transition is manifested by a decrease in both p34cdc2 kinase (MPF) and mitogen-activated protein (MAP) kinase activities, the onset of anaphase, CG exocytosis and the concomitant ZP modifications, and recruitment of maternal mRNAs. This time-dependent progression toward this interphase-like state may account for the increased susceptibility of such eggs to undergo parthenogenetic activation by decreasing the activity of critical enzymes involved in the maintenance of meiotic arrest (e.g., MPF and MAP kinase) to threshold levels that can readily be exceeded by a parthenogenetic stimulus.
Our previous study focused on these time-dependent changes in vivo; i.e., ovulated MII-arrested eggs were collected and analyzed at various times after hCG administration. Culture of MII eggs in vitro is widely used in many research and clinical applications. For example, many experimental designs involve the collection and culture of eggs prior to experimental manipulation. Likewise, recovery of eggs used for procedures of in vitro fertilization in humans, domestic animals, and endangered species frequently involve culture in vitro prior to insemination. This in vitro culture of eggs could compromise their subsequent fertilizability. In fact, in vitro culture results in changes in the distribution of cytoskeletal proteins in mouse [14] and human [15, 16] eggs, as well as in a reduction in zona solubility by proteases [17], a property indicative of ZP modifications [18, 19].
The purpose of this study was to test the hypothesis that the time-dependent progression toward an interphase-like state that is observed in vivo after hCG administration is exacerbated when ovulated eggs are harvested and then cultured in vitro. In these studies we assessed the effects of egg culture in vitro on 1) the onset of anaphase and H1 kinase and MAP kinase activities; 2) CG release and the accompanying ZP modifications; 3) the recruitment of maternal mRNAs; and 4) the ability to undergo ionophore-induced egg activation. The results of this investigation demonstrate fundamental differences between the responses of eggs to in vitro and in vivo environments and dictate certain aspects of the experimental design with the use of mouse eggs, as well as providing guidelines regarding the useful temporal window for clinical in vitro fertilization.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
At 13, 16, and 22 h post-hCG, MII-arrested eggs were removed from the oviducts of 8- to 10-wk-old female CF-1 mice (Charles River, Wilmington, MA; or Harlan, Indianapolis, IN) that had been injected i.p. with 5 IU of eCG (Sigma Chemical Co., St. Louis, MO), followed 48 h later by 5 IU of hCG. Eggs were collected in Whitten's medium [20] containing 0.01% polyvinyl alcohol (PVA) and 20 mM Hepes buffer, pH 7.4, at 37°C. Cumulus cells were removed using 0.01% hyaluronidase for 3–5 min in the same solution.
For in vitro-aged groups, after cumulus cell removal, eggs were either collected at 13 h and cultured to 16 h or 22 h post-hCG or collected at 16 h and cultured to 22 h post-hCG in either KSOM containing amino acids [21] or M16 [22]. At the end of culture, or for the in vivo groups immediately after cumulus removal, eggs were fixed in 3% paraformaldehyde in Dulbecco's PBS or prepared for analysis of H1 or MAP kinase activities as previously described [23] except for experiments in which eggs were treated with A23187, radiolabeled with [35S]methionine, or treated with 
-chymotrypsin.
Staining and Quantification of Egg CGs
ZPs were removed with 0.25% pronase in Earle's Hepes medium for 3–5 min, eggs were fixed as above, and CGs were stained with Lens culinaris agglutinin (LCA) coupled to biotin (Polysciences, Warrington, PA) and Texas Red streptavidin (Gibco BRL, Gaithersburg, MD), according to Cherr et al. [24] as modified by Ducibella et al. [25]; they were then visualized by fluorescence microscopy with a Nikon (Garden City, NY) x100 objective. Periodate-treated serum albumin [26] was used as a blocking reagent. Eggs were washed and mounted in glycerol:Dulbecco's PBS (1:1) with 0.02% sodium azide. The CG density in the cortex was determined by counting LCA-labeled CGs [27] and was expressed as the mean CG density per group. The CG-free domain normally present over the meiotic spindle [25] was not included in the CG density analyses. The area of the CG domain was calculated as follows. The CG-occupied domain normally composes 60% of the egg cortex [25]. The area of the cortex occupied by CGs was calculated by measuring the distance from the edge of that domain (in the vicinity of the egg equator) to the pole cortex, which is opposite to the egg chromatin. In a compressed egg, the ratio of this distance to the egg radius was converted to the area (calculations not shown).
Calculation of CG Release
For each group, the mean CG densities and corresponding areas of the CG-occupied domain were determined. The total mean number of CGs was calculated by multiplying the mean CG density and the mean CG-occupied domain area. Percentage CG release was calculated for each time point by comparing the in vivo- or in vitro-aged groups (IVA) to the 13-h in vivo group; i.e., % CG release = [1 - (IVA)/13 h] x 100. IVA is the product of the CG density and CG-domain area for the aged group, and 13 h is the control value likewise calculated.
Assay for ZP Hardening
The time-dependent dissolution of the ZP from the ZP-intact egg incubated in -chymotrypsin [28] was used to assess ZP hardening. Briefly, the eggs were pooled and treated with 1.0 mg/ml -chymotrypsin (type II, 40–60 U/mg protein; Sigma C-4129) in Earle's balanced salts solution with 0.3% polyvinylpyrrolidone at 37°C. Eggs were monitored every 5 min for up to 2 h following this treatment, and the time at which 50% of the eggs underwent complete ZP dissolution was assessed as the t50 for ZP dissolution.
Morphological Assessment of Meiotic Stage
The stage of egg meiotic maturation was determined from the chromatin configuration by costaining with 10–20 µg/ml each of 4,6-diamidino-2-phenyl indole and Hoechst 33258 for 10 min after LCA staining [29]. The separation of two sets of chromosomes and alignment toward the opposite poles were used as the criteria for anaphase. Decondensing chromatin and the formation of a polar body were used as an indication of telophase.
Histone H1 Kinase and MAP Kinase Assays
Histone H1 kinase activity (i.e., cdc2/cyclin B kinase activity) and MAP kinase activity in single eggs were measured as previously described using, respectively, histone H1 and a peptide substrate containing the MAP kinase consensus phosphorylation sequence found in myelin basic protein [23].
[35S]Methionine Radiolabeling of Eggs andTwo-Dimensional Gel Electrophoresis
For two-dimensional SDS-PAGE gels, eggs were radiolabeled in CZB medium [30] containing 1 mCi/ml of [35S]methionine (specific activity ~1500 Ci/mmol; Amersham, Arlington Heights, IL) for 2 h. Unfertilized or fertilized eggs were retrieved at 13 h post-hCG and labeled, or unfertilized eggs were retrieved at 13 h post-hCG and cultured for 8 h prior to labeling. The samples were then processed and subjected to two-dimensional gel electrophoresis using the Investigator 2-D Electrophoresis System (Millipore, Bedford, MA) as previously described [6]. Radiolabeled proteins were detected with a Molecular Dynamics (Sunnyvale, CA) PhosphorImager. Approximately 1.4 x 106 acid-precipitable cpm was applied to the gel, which was exposed for 3 days.
Treatment of Eggs with A23187
Eggs were treated with 5 µM A23187 for 5 min in Ca2+-free medium as previously described [4].
Statistical Comparisons
For treated and control groups, or for different times after hCG in vivo and in vitro, data were expressed as the mean value per group and compared statistically through use of the Student's unpaired t-test.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
We previously reported that eggs recovered as early as 16 h post-hCG displayed an increase in the incidence of the onset of anaphase, as well as a decrease in the activity of both MPF and MAP kinase, and that these changes occurred progressively with increasing times up to 22 h post-hCG [13]. In order to test the hypothesis that culture of eggs in vitro accelerates the time-dependent changes in these events, MII-arrested eggs harvested at 13 h post-hCG were cultured in either KSOM containing amino acids or in M16, two media that are frequently used for embryo culture. We noted in these studies no significant statistical differences between results with these two media, and only the results with KSOM are reported except for those in Figure 1.
|
In contrast to the time-dependent entry of ~10% of the eggs into anaphase observed 16 h post-hCG in vivo (Fig. 1b, Table 1, [13]), none of the eggs removed from the oviduct at 13 h and cultured to either 16 h or 22 h post-hCG exhibited an anaphase chromatin configuration; essentially all of the eggs displayed an MII appearance with chromosomes tightly aligned on the metaphase plate (Fig. 1a and Table 1). When ovulated eggs were harvested at 16 h post-hCG and then cultured to a time that corresponded to 22 h post-hCG, i.e., cultured for 6 h, a small percentage of eggs exhibited a telophase chromatin configuration or the formation of a pronucleus (Fig. 1c and Table 1). In contrast, the percentage of eggs that entered anaphase by 22 h post-hCG in vivo increased 3- to 4-fold, but none of these eggs exhibited pronucleus formation (Table 1). Similarly, the time-dependent decreases in histone H1 kinase and MAP kinase activities that occur in vivo after hCG administration were not observed when the eggs were harvested at 13 h post-hCG and cultured to a time that corresponded to either 16 or 22 h post-hCG, or when the eggs were harvested at 16 h post-hCG and cultured to a time that corresponded to 22 h post-hCG (Fig. 2). Thus, in contrast to the anticipated result that culture would promote meiotic resumption, culture of eggs in vitro in fact retarded and/or prevented this transition.
|
|
Effect of Culture on the Time-Dependent Changes in CG Exocytosis, ZP Modification, Recruitment of Maternal mRNAs, and Egg Activation
The unanticipated observation that culture inhibited the progression toward the interphase-like state suggested that the other events of egg activation, i.e., CG exocytosis, the ZP modifications, and maternal mRNA recruitment, would also be prevented from occurring during the in vitro culture period. In fact, such was the case. For example, as we previously demonstrated [13], eggs isolated at later different times after hCG administration (i.e., 16 and 22 h) displayed an increased incidence of entry into anaphase that was associated with an increase in CG loss as compared to that in eggs isolated at 13 h post-hCG (Table 2). In contrast, eggs collected and cultured in vitro for various periods of time revealed a reduced incidence of meiotic resumption, as well as a reduced CG loss, in comparison to ovulated eggs retrieved at parallel times (Table 2). For example, all eggs cultured from 13 to 22 h remained at MII with a minimal CG release (~3%), while over a third of eggs at 22 h in vivo progressed to anaphase with a mean CG release of 49%. Also, CG release in eggs cultured from 16 to 22 h (13.9%) was similar to that in eggs at 16 h in vivo (11.8%), indicating that no additional CG release occurred in vitro (Table 2). It should be noted that <5% of cultured eggs did resume meiosis and that these eggs also exhibited a release of CGs (Table 2).
|
Consistent with the inhibitory effect of culture on CG release was a corresponding inhibitory effect on the t50 for the chymotrypsin-induced dissolution of the ZP, which is diagnostic for the ZP hardening [31] that occurs following CG exocytosis [27] (Fig. 3). Eggs isolated at various times after hCG administration displayed a time-dependent increase in the t50 for ZP dissolution as we previously reported [13], and this is consistent with the time-dependent progression toward the interphase-like state and increase in CG release (Table 2). In contrast, eggs collected at 13 h and cultured in vitro to 16 h or to 22 h demonstrated a reduced t50 for ZP dissolution in comparison to eggs at 16 h or 22 h in vivo, respectively (Fig. 3). Just as CG release increased with longer duration in vivo, the t50 increased similarly; e.g., eggs cultured from 16 to 22 h had a higher t50 than those cultured from 13 to 22 h. These results were consistent with inhibition of CG release and inhibition of progression toward the interphase-like state (Table 2).
|
We previously demonstrated that progression toward the interphase-like state was also accompanied by recruitment of maternal mRNAs that are normally translated after fertilization or egg activation [13]. Since culture did not support progression into the interphase-like state, as anticipated these changes in mRNA recruitment did not occur (Fig. 4).
|
The increased incidence of entry into the interphase-like state that occurs in vivo after recovery of eggs at various times following hCG administration was accompanied by an increase in the susceptibility of these eggs to be activated by parthenogenetic agents (Fig. 5, [13]). This time-dependent increase was not observed when the eggs were harvested 13 h post-hCG and then cultured for either 3 or 9 h prior to ionophore treatment or were collected 16 h post-hCG and then cultured for 6 h prior to ionophore treatment (Fig. 5).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Maintenance of arrest in MII requires high levels of p34cdc2 and MAP kinase activities, and entry into interphase requires a decrease in these activities below a critical threshold [23, 32]. Our observation that these activities remain elevated in eggs that are cultured in vitro likely accounts for the maintenance of arrest in MII. This contrasts with the decrease in these activities that occurs in a time-dependent fashion while the eggs reside in the oviduct [13]. This time-dependent decrease likely accounts for the progression toward an interphase-like state of eggs that reside in the oviduct, leading to the aforementioned events of egg activation. In contrast, cultured eggs do not show such a pronounced decrease in p34cdc2 and MAP kinase activities and hence have a cytoplasm that maintains meiotic arrest. This cytoplasmic environment is likely not compatible with other events associated with the progression toward interphase, e.g., CG exocytosis and recruitment of maternal mRNA.
The maintenance of elevated levels of p34cdc2 and MAP kinase activities in the cultured eggs may also account for their reduced ability to be activated by calcium ionophore in contrast to eggs that reside in the oviduct for a comparable period of time [13]. Fertilization of mammalian eggs is accompanied by an initial calcium transient that is followed by repetitive calcium oscillations. Although the exact function of these calcium oscillations is not clear, it is most likely that they play a role in exit from M phase by driving p34cdc2 activity below the threshold required for maintenance of meiotic arrest [13, 33, 34]. These calcium oscillations could, in principle, sustain the activation of the calcium-dependent pathway; this pathway leads to a decrease in MPF activity in the absence of tonically high intracellular calcium concentrations that could be injurious to the fertilized egg. In the eggs cultured in vitro, the single calcium transient induced by ionophore may be insufficient to drive the p34cdc2 and MAP kinase activities below the critical threshold level needed for entry into interphase that maintains elevated levels of this activity. This is in contrast to observations in eggs that reside in the oviduct for extended periods of time in which p34cdc2 and MAP kinase activities show a more pronounced decrease. Thus, the single calcium transient induced by ionophore treatment may be sufficient in this case to drive these activities to below threshold levels. As previously discussed [13], the failure of ionophore to activate freshly ovulated eggs possessing high levels of p34cdc2 and MAP kinase activities may be due to the inability of the single calcium transient to drive these activities below the critical threshold required for entry into interphase.
A second important difference between eggs aged in vivo and in vitro to 22 h post-hCG was the status of the CGs. Unlike eggs in vivo, a large percentage of which undergo both anaphase onset and CG exocytosis [13], those residing in vitro from 13 to 22 h post-hCG underwent neither anaphase onset nor significant CG release. While MII-arrested eggs collected at 16 h post-hCG display a low level of CG release, no further release is observed when these eggs are further cultured in vitro for an additional 6 h (i.e., to a time that corresponds to 22 h post-hCG). The failure to detect a significant CG release is consistent with the reduced t50 for ZP digestion in eggs cultured from 13 to 22 h compared to that for eggs in vivo to 22 h post-hCG. As expected, those eggs residing for longer times in vivo (13–16 h) and subsequently cultured from 16 to 22 h have longer t50s, approaching that of the 22-h in vivo group. It should be noted, however, that some of these cultured eggs do enter anaphase and that these eggs also undergo a notable, but variable, degree of CG release.
Although the molecular basis for the observed differences in spontaneous activation events occurring in vivo compared to those occurring in vitro remains unexplained, it is possible that differences in ionic composition of the in vivo and in vitro environments may affect cell cycle progression. A comparison of the composition of the culture media and tubal fluid reveals that the concentrations of ions in culture media are similar to those determined in both mouse [35] and human [36] tubal fluid. In fact, embryo culture media were largely based on these calculated values. A notable exception is the potassium concentration that is 5- to 10-fold higher in mouse tubal fluid than in M16 and KSOM. In the hamster egg, fertilization results in elevations in intracellular calcium and membrane hyperpolarization associated with increases in the permeability of the membrane exclusively to potassium ions [37]. The higher potassium concentration present in mouse tubal fluid as compared to culture medium is likely to increase the egg's resting plasma membrane potential. This higher membrane potential in vivo could allow the opening of a small percentage of voltage-dependent calcium ion channels [38]. This could result in a small influx of calcium that could account for the observed anaphase onset without further cell cycle progression. Conversely, the lower resting membrane potential in vitro would prevent this low level of calcium influx. This may not be the case, however, since the resting potential has been measured in vitro at about -40 mV [39], where most channels are not open. It should be noted that the membrane potential and subsequent status of ion channels in the in vivo milieu of the oviduct are not known for mammalian eggs.
It is also possible that the aforementioned differences may be attributed to specific non-ionic factors in vivo that actively promote the entry into anaphase in ovulated eggs. For example, oviduct fluid has specific glycoproteins synthesized at the time of estrus [40, 41] and a higher protein concentration (10.9 mg/ml [42]) than KSOM or M16. Putative factors could come from either the oviduct epithelium or cumulus cells. An alternate hypothesis is that conditions in vitro may inhibit anaphase onset; e.g., mouse embryos undergo fewer cell divisions in vitro than in vivo.
Our results also have implications for the time at which mouse eggs should be harvested after hCG administration for the purposes of conducting experiments aimed at understanding the molecular basis of fertilization. We suggest that eggs be harvested at a time shortly after ovulation (between 13 and 16 h post-hCG) in order to ensure that the eggs remain in MII, since at later times the eggs are starting progress toward an interphase-like state [13]. Although placing eggs in culture apparently extends significantly the time of MII arrest, nevertheless the experimental window is restricted, since cumulus-free eggs cultured in vitro display a partial ZP hardening in a time-dependent manner (this study, [17]). It is also not known how long eggs arrested in MII can be cultured (prior to fertilization) while maintaining competence to undergo normal development. The present study examines only a subset of egg physiological parameters.
Although the eggs used in the studies reported here were cultured in vitro in the absence of cumulus cells (which is in contrast to most clinical in vitro fertilization protocols), our results may have implications for human in vitro fertilization. Our finding that in vitro culture reduces the incidence of spontaneous egg activation, which would negatively impact normal fertilization and development [13, 43], suggests that mature MII eggs should be removed and placed into an appropriate culture medium soon after the completion of meiotic maturation or ovulation. This would avoid adverse effects of intrafollicular aging [44] or tubal spontaneous activation. In this regard, cultured human eggs that fail to fertilize in in vitro fertilization or intracytoplasmic sperm injection have a very low incidence of spontaneous anaphase onset after prolonged culture [45, 46]. Nevertheless, prolonged culture is not advisable, since spontaneous CG release [46, 47] leading to ZP hardening can occur, and this would result in reduced fertilization rates. Consistent with our results with the use of mouse eggs is the previous finding that one half of human eggs that fail to fertilize are found to have considerable ZP2 conversion a day after insemination (although the timing of conversion was not determined for clinical and ethical reasons) [46].
Optimizing the conditions that maintain the MII egg's normal competence to undergo fertilization and development would be advantageous for both experimental studies and clinical applications. It remains to be demonstrated whether the in vitro environment actively inhibits a tendency for eggs to undergo spontaneous activation events or whether the in vivo environment promotes activation events in an otherwise quiescent, arrested egg.
|
FOOTNOTES |
|---|
2 Correspondence: Richard M. Schultz, Department of Biology, University of Pennsylvania, 415 South University Avenue, Philadelphia, PA 19104–6018. FAX: 215 898 8780; rschultz{at}mail.sas.upenn.edu
Accepted: July 31, 1998.
Received: June 24, 1998.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  A. J Gardner, C. J Williams, and J. P Evans Establishment of the mammalian membrane block to polyspermy: evidence for calcium-dependent and -independent regulation Reproduction, February 1, 2007; 133(2): 383 - 393. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Tatone, M. C. Carbone, R. Gallo, S. Delle Monache, M. Di Cola, E. Alesse, and F. Amicarelli Age-Associated Changes in Mouse Oocytes During Postovulatory In Vitro Culture: Possible Role for Meiotic Kinases and Survival Factor BCL2 Biol Reprod, February 1, 2006; 74(2): 395 - 402. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y.-L. Miao, X.-Y. Liu, T.-W. Qiao, D.-Q. Miao, M.-J. Luo, and J.-H. Tan Cumulus Cells Accelerate Aging of Mouse Oocytes Biol Reprod, November 1, 2005; 73(5): 1025 - 1031. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Alikani, T. Schimmel, and S. M. Willadsen Cytoplasmic fragmentation in activated eggs occurs in the cytokinetic phase of the cell cycle, in lieu of normal cytokinesis, and in response to cytoskeletal disorder Mol. Hum. Reprod., May 1, 2005; 11(5): 335 - 344. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. B. Wortzman and J. P. Evans Membrane and cortical abnormalities in post-ovulatory aged eggs: analysis of fertilizability and establishment of the membrane block to polyspermy Mol. Hum. Reprod., January 1, 2005; 11(1): 1 - 9. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. J. Tunquist and J. L. Maller Under arrest: cytostatic factor (CSF)-mediated metaphase arrest in vertebrate eggs Genes & Dev., March 15, 2003; 17(6): 683 - 710. [Full Text] [PDF]
|
|
|
|
|
|
A. C. Gordo, P. Rodrigues, M. Kurokawa, T. Jellerette, G. E. Exley, C. Warner, and R. Fissore Intracellular Calcium Oscillations Signal Apoptosis Rather than Activation in In Vitro Aged Mouse Eggs Biol Reprod, June 1, 2002; 66(6): 1828 - 1837. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. B.Mailhes, D. Young, G. Caldito, and S.N. London Sensitivity of mouse oocytes to nicotine-induced perturbations during oocyte meiotic maturation and aneuploidy in vivo and in vitro Mol. Hum. Reprod., March 1, 2000; 6(3): 232 - 237. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. J. Tarín, S. Pérez-Albalá, A. Aguilar, J. Miñarro, C. Hermenegildo, and A. Cano Long-Term Effects of Postovulatory Aging of Mouse Oocytes on Offspring: A Two-Generational Study Biol Reprod, November 1, 1999; 61(5): 1347 - 1355. [Abstract] [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
23 eggs per group. Data are expressed as the mean ± SEM. Means with different letters indicate statistically significant differences (p < 0.05).
