Alteration of mature neuronal marker of β-III tubulin expression in differentiated SH-SY5Y cells by refinement of foetal bovine serum concentration

Background The process of differentiating neuroblastoma cells (SH-SY5Y) is crucial for obtaining mature neuronal markers. However, there is variability in the concentrations of foetal bovine serum (FBS) used in the differentiation media, ranging from 1 to 10%. This inconsistency in FBS concentrations may contribute to the inconsistent differentiation of cells. To improve the utility of the SH-SY5Y cell line as a model for neuronal cell culture, we investigated the impact of FBS concentrations in the differentiation media using Dulbecco’s modified eagle medium and 10 μM all-trans-retinoic acid (ATRA). The aim of this study was to optimise the concentrations of FBS in the differentiation media of SH-SY5Y cells. Our study focused on assessing the length of dendrites in neuronal cells and the expression of β-III tubulin, a marker indicative of mature neurons. SH-SY5Y cells were differentiated with 10 µM of ATRA for 7 days. Four treatment groups with different FBS concentrations (1%, 3%, 5%, and 10%) were examined to assess the alteration of cellular morphology and dendritic length. The expression of the mature neuron marker β-III tubulin was evaluated using immunocytochemistry technique. Results SH-SY5Y cells’ dendrite length was significantly longer ( p < 0.05) when there was a higher concentration of FBS in the differentiation medium. The result was confirmed with the significant increase of β-III tubulin expression ( p < 0.001) of the differentiated SH-SY5Y cells that have been incubated with higher concentrations of FBS in the differentiation medium. Conclusions We concluded that optimised concentrations of FBS in the differentiation media display longer length of the dendrites and express higher production of β-III tubulin in the differentiated SH-SY5Y cells. The consistency of serum concentration used in the differentiation media is important to produce a sustainable in vitro neuronal model of SH-SY5Y cells for neurodegenerative studies.


Background
The production of in vitro cultures of neuronal cells is particularly difficult because of the inability of mature neurons to proliferate [1][2][3].In order to progress further towards a better understanding of the functions of the nervous system, we need to be able to produce in vitro cultures of neuronal cells.The establishment of an immortalised secondary cell line derived from neuronal tumours is one of the approaches to overcome these challenges [2].
Several in vitro studies particularly involving cells derived from primary tissue [4],stem cells such as human induced pluripotent stem cells (iPSC) [5], human olfactory bulb neural stem cells (hOBNSCs) [6], and immortalised cell lines [7] have contributed significantly to enhance our understanding of neurodegenerative diseases.However, the immortalised cells such as neuroblastoma (SH-SY5Y) are relatively easy to maintain, providing unlimited cell numbers and exhibiting minimum variation between cultures.These cell lines could show physiological differences due to their cell's origin from where they were derived.Therefore, these cell lines will be induced to exhibit a neuronal phenotype by manipulating the culture conditions through specific differentiating agents or specific growth factors [2].
β-III tubulin belongs to the beta-tubulin protein family, specifically as a class III member.These beta tubulins constitute essential elements in constructing the microtubule network within mammals.Unlike general tubulins that participate in various cellular functions such as mitosis and motility, β-III tubulin is exclusively found in neurons [15].Its presence is particularly significant during the initial stages of neuronal differentiation [16], making it a reliable indicator for mature neurons.As a result, β-III tubulin is now recognised as a definitive marker for mature neurons [17].
Several agents like retinoic acid, phorbol esters, and dibutyryl cAMP can be used to induce proliferation and differentiation in SH-SY5Y.The most common differentiating agent used is retinoic acid with a standard concentration of 10 μM [7,18].Several protocols on differentiating the SH-SY5Y cell line using retinoic acid have been established.However, the percentages of FBS used in the differentiation media vary from 1 to 10%.Despite the extensive use of differentiated SH-SY5Y cells in neurodegenerative research, it is still unclear how the different serum concentrations in the differentiation media influence the cells during the differentiation process.
Previous studies showed that SH-SY5Y is sensitive to FBS withdrawal [19,20].Media composition for embryonic stem cells can also influence the regulation of pluripotency-associated gene expression of the cells [21].This suggests that different percentages of FBS might influence the outcome of the differentiation.Besides, one of the important considerations for in vitro studies is the ability of SH-SY5Y cells to differentiate into neuron-like phenotype that expresses a mature neuronal marker [7].In this connection, we aim to investigate the effect of different percentages of FBS towards the morphology of differentiated cells in terms of dendritic length as well as in the expression and the distribution of mature neuronal marker of β-III tubulin.

Cell culture and differentiation
SH-SY5Y cells from American Type Culture Collection (ATCC) were grown in DMEM medium, which was enriched with 15% foetal bovine serum, 50 U/ml penicillin, and 50 μg/ml streptomycin (all sourced from Gibco, Paisley, UK).The cell cultures were maintained at 37 °C in a humidified atmosphere with 5% CO 2 .The culture medium was refreshed every 3 days, and cells were subcultured when they reached approximately 80% confluence.For the differentiation process, the cells were preconditioned with 10 μM all-trans retinoic acid (ATRA) (MPBiomedicals, LLC) for 1 week.The differentiation medium, as outlined in Table 1, was replaced every 3 days.

Morphological classification
The cells underwent differentiation for 7 days.Subsequently, the morphology of live cells was observed on both day 0 and day 7 using an Axiovert 25 phase-contrast microscope from Carl Zeiss, equipped with a Media Cybernatics Digital camera and Image Pro-express software (Media Cybernatics, USA).Three specific characteristics (Table 2) were examined by at least 3 examiners, including growth inhibition, length of neurite elongation, and neurite branching.The morphological classification was conducted according to the criteria outlined by Agholme et al. [1].Growth inhibition was graded by comparing the cell count on Day 0 with that on Day 7.
The experiments were conducted in triplicate.

Measurement of dendrite length
Neurites were traced from phase-contrast images of each treatment group using Simple Neurite Tracer from ImageJ.A hundred neurites were measured from each group of treatment.

Immunocytochemistry
Cells were seeded in 4-well chamber slides coated with poly-D-lysine (Corning, USA), with a seeding density of 100,000 cells per well.The fixation and permeabilization steps followed the provided protocol of the Image-iT ® Fixation/Permeabilization Kit.Subsequently, the samples underwent overnight incubation at 4 °C with the primary antibody anti-neuron-specific β-III tubulin (Cat#: ab229590) at a dilution of 1:200 (Abcam, Cambridge, UK) in the blocking solution provided by the kit.
After washing with Wash buffer, the cells were incubated with secondary Alexa Fluor-488 goat anti-rabbit antibodies (Cat#: ab150077) diluted 1:200 in Wash buffer (Abcam, Cambridge, UK) at room temperature for a 1-h incubation period.Finally, the samples were mounted using Fluoroshield Mounting Medium with 4' ,6-diamidino-2-phenylindole (DAPI) (Cat#: ab104139) and left to incubate overnight (Abcam, Cambridge, UK).Cell examination was conducted using an Olympus BX41 microscope equipped with two filters at 360 and 488 nm, in conjunction with the analysis software from Olympus Corporation.The intensity of neuron-specific β-III tubulin was assessed using ImageJ software (NIH).For every treatment, three images were analysed.The intensity of the selected area was measured, and the overall mean of neuron-specific β-III tubulin intensity was calculated (n = 3).

Statistical analysis
The data were analysed using one-way analysis of variance (ANOVA) and followed by a Tukey post hoc test for comparison between treatment groups.The results were presented as mean ± standard deviation (SD).Spearman's correlation test was used to analyse the correlation between the mean of dendritic length and intensity of β-III tubulin.A p value < 0.05 was considered statistically significant.The SPSS ® 2.0 package (IBM ® SPSS ® ) was used for all statistical calculations.

Morphological changes
Most SH-SY5Y cells displayed morphological alterations within a short period of 3-5 days.Initially, the undifferentiated cells were presented as clusters of rounded cells with short neurites.Following treatment with ATRA, these cells became elongated, exhibiting longer neurites compared to their undifferentiated counterparts.Additionally, ATRA-treated cells formed less clustered arrangements.Cells subjected to 10% FBS differentiation media demonstrated longer dendrite length compared to other groups (Fig. 1).To assess morphological distinctions associated with varying FBS percentages in SH-SY5Y differentiation media, morphological classification was conducted based on Agholme et al. [1].However, there were no discernible differences observed in  terms of growth inhibition, length of neurite elongation, as well as neurite branching across all stimulation groups (Table 3).Dendrite lengths were meticulously measured and compared as part of supplementary analysis for each group.
The longest dendrite length was found in differentiated cells stimulated with 10% FBS differentiation media (Fig. 1).The results of a one-way ANOVA analysis indicated a statistically significant distinction among groups (F (3, 399) = 9.97, p < 0.001) (Table 4).Tukey post hoc testing demonstrated a significant difference in dendrite length between cells that were stimulated with 10% FBS-DM and 1% FBS-DM, as well as between 10% FBS-DM and 3% FBS-DM (p < 0.05) (Table 4).

Expression of mature neuronal marker of β-III tubulin
The intriguing findings were further investigated concerning the mature neuronal marker β-III tubulin using immunocytochemistry.In undifferentiated cells, β-III tubulin expression was observed in the cell cytoplasm.Upon differentiation, there was a noticeable shift, with β-III tubulin now distributed along the dendrites, exhibiting a more neuritic pattern.The quantification of β-III tubulin expression revealed a significant increase in differentiated cells treated with 10% FBS-containing media (Fig. 2).
Statistical analysis using one-way ANOVA demonstrated a highly significant difference between groups (F (4; 854.02) = 185.77,p < 0.001) (Table 5).Further analysis through Tukey post hoc testing indicated significant differences in β-III tubulin intensity (P < 0.05 as determined by Tukey's post-hoc test) across all paired groups, except for the comparison between 1% FBS, DM, and 3% FBS, DM differentiated cells (Table 5).The correlation between the mean of dendritic length observation and β-III tubulin intensity was analysed using Spearman's rank correlation test.There was a significant, strong, and positive relationship between mean of dendritic length and β-III tubulin intensity in all groups (p < 0.01, r = 0.86) (Fig. 3).

Discussion
In accordance with previous observations, the differentiation of neuroblastoma SH-SY5Y cells resulted in the development of neuron-like cells with morphological alterations such as an increase in neurite length within 3-5 days [9,22] of differentiation.In our study, there were no notable distinctions noted in the morphological classification of SH-SY5Y cell line post-differentiation.However, upon closer examination focusing on dendrite length, it was suggested a contrary outcome.Morphologically, a specialised neuron cell is characterised by possessing a neurite length surpassing that of the cell body, typically averaging greater than 10 μm in length [23].According to the results, all treatment groups had dendrite length of more than 10 µm.This indicates that all stimulated groups had successfully differentiated after treatment for 7 days.Our results were consistent with previous studies which reported that SH-SY5Y cell line should be differentiated for at least 7 days to become mature neurons [12,24].Our results showed that SH-SY5Y cells that had been differentiated with 10% FBS and 10 μM ATRA had the longest neurite in comparison to other treatment groups.
In undifferentiated cells, the expression of β-III tubulin was observed to be distributed in the cytoplasm of the cells.Upon differentiation, β-III tubulin is distributed along the dendrite and becomes more neuritic.From the results, although the cells were morphologically differentiated (having a dendrite length of more than 10 µm), the cells did not necessarily express a mature neuronal marker of β-III tubulin in the neurites.The higher the percentage of FBS in the DM, the greater the expression of β-III tubulin in the neurite, and the higher the intensity of β-III tubulin in the differentiated cells will be.In  this case, the expression of β-III tubulin was the highest in differentiated cells with 10% FBS DM.
The key process in early neuronal differentiation is the sprouting of neurites, which will subsequently develop into axons and dendrites [25].β-III tubulin expression is associated with early neuritogenesis [15].In a theoretical context, the division of cells containing tubulin specific to neurons within their mitotic spindles resulted in a decrease in cell proliferation.This process led to the emergence of postmitotic neurons with the tubulin evenly distributed throughout cellular processes [26].Earlier investigations proposed an association between the expression of β-III tubulin in mature neurons and the  presence of exceptionally stable microtubules, playing pivotal roles in both the development and maintenance of neurites [27][28][29][30].Furthermore, it is conjectured that β-III tubulin may contribute to the extension of neurites by promoting microtubule polymerization in the early phases of neuritic development in neuroblastoma cells [31].
The results of this study suggest the availability of nutrients in the differentiation media may influence the expression of β-III tubulin.An increase in nutrient availability (percentage of FBS in the differentiation media) significantly influences the expression of β-III tubulin, thereby leading to longer dendrite length formation.The serum is widely employed in culture media, serving as a source of essential hormonal factors that promote cell division, development, and proliferation, along with facilitating differentiation activities.Additionally, it also contains transport proteins responsible for carrying hormones, such as transcortin, as well as minerals, trace elements, and lipids.Moreover, serum contributes attachment and spreading factors similar to those found in the extracellular matrix.It plays a role in maintaining pH and inhibiting proteases, either directly or indirectly, by providing stabilizing and detoxifying factors [32,33].The findings of this study may also indicate that the differentiation of SH-SY5Y could be influenced by the availability of nutrients or the percentage of FBS in the differentiation media.
The viability of SH-SY5Y cells was found to be influenced by serum reduction, as demonstrated in prior research [3,20].Another study revealed that withdrawing serum could elevate monoamine oxidase (MAO) activity, increase Caspase 3 activity, and induce apoptosis [19].Elevated Caspase 3 levels are linked to reduced neurite outgrowth [34,35].These findings may elucidate the varied results observed in differentiated SH-SY5Y cells when using different percentages in the differentiation media.
Several methods have been outlined to induce the maturation of this cell line into fully developed neurons [1,3,9,11,12,36,37].The concentration of serum used in the DM ranges from 1 to 10%.This study has demonstrated that varying serum percentages lead to distinct results in both dendrite length and the expression of the mature neuronal marker β-III tubulin in differentiated SH-SY5Y cells.
The SH-SY5Y culture comprises of two distinct cell populations: the neuronal-like, retinoic acid-sensitive 'N' subtype, and the substrate-adherent, differentiationresistant 'S' subtype [38].Several of the SH-SY5Y cell line's differentiation procedures entail serum reduction to inhibit the growth of the 'S' subtype, as its proliferation may outpace that of the 'N' subtype.In our experiment, we noticed a minimal presence of the 'S' subtype in our differentiated cell population.To reduce the 'S' subtype, we implemented a strict schedule for passaging the SH-SY5Y cell line, utilizing only the adherent population.Our observations revealed that cells with a passage number exceeding 15 tended to die before completing the differentiation process.Typically, cells would start extending neurites, but by day 3, they exhibited signs of poor health and eventually perished.Hence, our differentiation protocol employed cells with a passage number less than 15.Neurite elongation usually begins, but by day 3, unhealthy cells appeared and eventually died.Therefore, our differentiation protocol utilised cells with a passage of less than 15.Recent findings recommend an optimal differentiation medium for studying Parkinson's disease with SH-SY5Y, involving 3% serum in retinoic acid [39].Our study aims to optimise the differentiation medium for studying Alzheimer's disease (AD) with SH-SY5Y.Variations in results stem from different methodologies and measured markers.

Limitations
Several limitations were encountered during the experiment.The SH-SY5Y cell line showed inherent variability in response to ATRA, which is a fragile compound that easily oxidises and is very sensitive to light exposure.The small sample size made it difficult to measure meaningful differences between the stimulation groups, limiting the generalizability of our findings.Additionally, we were unable to conduct the qPCR and western blot due to resource constraints.However, the results of immunofluorescence of β-III tubulin expression from our study demonstrated the specific localization of β-III tubulin, which is an important marker for mature neurons during the initial stage of neuron differentiation.

Conclusion
This study showed that the serum plays a crucial role in determining the outcomes of SH-SY5Y cell line differentiation.Therefore, it is important to understand that the serum concentration in the differentiation media will have an impact on SH-SY5Y cell differentiation.Our findings suggested that a protocol incorporating a higher percentage of FBS during the differentiation of SH-SY5Y cell line generates a sustainable population of cells exhibiting a more mature, human-like neuronal morphology.
Furthermore, our research reveals that undifferentiated SH-SY5Y cells lack certain neuronal characteristics.Hence, using the differentiated SH-SY5Y cells is recommended for a more accurate representation of neurons in Alzheimer's disease studies.Further research to explore the specific properties of serum that impact neuritogenesis or differentiation in SH-SY5Y cells could unveil a new pathway for understanding how nutrients affect the neuritogenesis of neuronal cells.

Fig. 1
Fig.1Morphology of A undifferentiated SH-SY5Y and after 7-day differentiation under stimulation of B 1%, C 3%, D 5%, and E 10% FBS ATRA-media under ×200 total magnification.The higher percentage of FBS in the differentiation media leads to longer dendrite length between adjacent networking cells (yellow arrowhead)

Fig. 2
Fig. 2 Immunofluorescence of β-III tubulin under ×200 total magnification.A higher percentage of FBS leads to significantly greater intensity of β-III tubulin in differentiated SH-SY5Y cells

Table 1
Differentiation medium used for different treatment groups

Table 2
Post-treatment cells criteria are ranked based on cell growth coverage and neurites characteristics

Table 3
The morphological features graded based upon the levels of growth inhibition, length of neurite elongation and neurite branching

Table 4
Mean of dendritic length in different stimulation groups a One-Way ANOVA test b Dendrite length between differentiated cells stimulated with 10% FBS, DM and 1% FBS, DM, 10% FBS, DM and 3% FBS, DM were significantly different at p < 0.05 by post-hoc test Tukey's procedures

Table 5
The intensity of β-III tubulin in different stimulation groups a One-Way ANOVA testbThe intensity of β-III tubulin is significantly different (p < 0.05 by post-hoc test Tukey's procedures) among all paired groups except between 1% FBS, DM and 3% FBS, DM differentiated cells