The composition of tooth enamel and dentin, including the abundance of proteins, is critical for dental health and tooth properties (1–3). The high mineral content in tooth enamel confers remarkable durability to teeth. It also preserves the incremental growth structures and entraps the ephemeral organic matrix, which mediates biomineralization (4–6). With advancements in analytical methods, teeth and tooth enamel are appreciated in scientific research as archives of biological information about development, environment, identity, behavior, as well as revealing prenatal and postnatal exposures, determining sex, and capturing snapshots of the proteome at the time of tooth development (7–14). In addition, the analysis of the organic portion in tooth enamel is key to understanding the causes and mechanisms of dental defects and to developing strategies to improve tooth enamel properties and resistance to decay (15, 16). However, the reliability of information derived from the proteomic analyses of the organic material within enamel greatly depends on its preservation under ex vivo conditions until the time of analysis.

The effects of storage conditions on the chemical composition and physical properties of dental enamel have previously been demonstrated to alter its mechanical properties (17–20). It has also been shown that modifying the protein matrix of enamel significantly affects the outcomes of biomaterial studies aimed at enhancing restorative outcomes (21, 22). However, no systematic comparisons have been published on how storage conditions affect the proteome of enamel.

In this study, we investigate how common storage conditions for extracted teeth affect the outcomes of enamel proteomic analyses. Addressing this question is essential for establishing standardized and reliable protocols for storing extracted or exfoliated teeth in future studies. Furthermore, understanding of how specific storage conditions affect the preservation of different protein families is critical for identifying potential biases in proteomic datasets and guiding protocol development for studies targeting specific proteins.

The number of identified proteins is highest in EtOH, similar in −80°C conditions, and less than half in Air and PBS. We identified 454, 460, 232, and 221 proteins in the conditions of −80°C, EtOH, Air, and PBS, respectively (Figure 2A). The number of unique proteins identified only in a particular condition was 77 for storage at −80°C and 60 for EtOH samples, with 177 common proteins in these experimental groups that were not identified in the PBS or Air groups. In contrast, PBS and Air samples had 33 and 31 unique proteins, respectively. Based on the Shannon diversity index (Figure 2B), storing enamel at −80°C resulted in the highest diversity of the extracted proteome, while the diversity of the proteome was lower in EtOH and PBS stored samples, albeit similar in those two groups. In contrast, the highest homogeny between replicates was seen in Air stored samples.

The relative abundance of enamel-specific proteins differs between storage conditions. We identified enamel-specific proteins in all samples, including amelogenin (AMEL), ameloblastin (AMBN), enamelin (ENAM), and odontogenesis-associated phosphoprotein (ODAPH) (Figure 2C) with the following relative abundances: AMEL was highest in PBS and lowest in −80°C; AMBN was highest in PBS and lowest in −80°C; ENAM was highest in PBS and lowest in EtOH; and ODAPH was highest in EtOH and lowest in PBS. The enamel proteases kallikrein-related peptidase-4 (KLK4) and matrix metalloproteinase-20 (MMP20) were also identified and quantified. The highest relative abundance of KLK4 was in PBS and the lowest was in EtOH and equal in EtOH and Air, while the highest relative abundance of MMP20 was in Air and was lowest when stored at −80°C (Figure 2C).

We also identified non-enamel-specific proteins that are usually detected in enamel proteomic studies and considered biologically relevant. Representing collagen, Collagen 1 alpha chain was identified in all samples; it had the highest relative abundance when stored air-dried and the lowest abundance when stored at −80°C (Figure 2C). Serum albumin (ALB) had the highest abundance when the samples were stored in PBS, while the lowest abundance was observed when they were stored in Air (Figure 2C). All proteins and their normalized relative abundances are listed in Supplementary Sheet 1. Proteins that are significantly different across different storage conditions can be seen in Supplementary Figure S2 and Sheet 1.

The order of protein classes, if ranked by the highest number of proteins, differs between samples and their storage conditions. A comparison of protein classes with the highest number of proteins between samples of different storage conditions is shown in Figures 3A–D (29). The top protein class was metabolite interconversion enzyme (PC00262) in all storage conditions. The closest match between the distribution of the top 10 protein classes was between −80°C and EtOH conditions (Figures 3A,B); however, the following differences remained: defense/immunity class (PC00090) was missing in the top 10 in EtOH, while this is the top ninth category in the −80°C group; the extracellular matrix protein group (PC00102) was the seventh category in EtOH-stored samples but was the 10th protein category in −80°C samples. Interestingly, the category of extracellular matrix proteins was the second-highest category in the Air samples (number sign in Figure 3C) and ranked ninth in PBS storage conditions (Figure 3D).

Some protein classes were more conserved in EtOH and PBS. The protein-modifying enzyme group (PC00260) was the second-highest category in PBS samples (asterisk in Figure 3D), while it was the fifth highest category in the other three conditions. The RNA metabolism protein (PC00031) category was the third highest protein category among the −80°C and EtOH groups, but it was not among the top 10 in the Air and PBS groups. Finally, the individual proteins and their relative abundances are shown in Figure 3E. Each protein is represented as a line, and the protein classes that they belong to are indicated with their color on the left and right Y axis (Figure 3E).

Various methods are used to store extracted teeth before analysis, depending on the availability of resources in a clinical setting and the practicality of sample storage and transport. For dental educational purposes, where no downstream chemical analyses are planned, the CDC recommends autoclaving or immersion in 10% formalin for 2 weeks (30). However, these methods significantly alter the mineral and organic components of the tooth. If proteins are to be studied, autoclaving will not only destroy infectious agents but also alter the organic matrix of the dental tissue, and formalin fixation will result in the cross-linking of molecules, affecting protein extraction from tooth samples. Therefore, less aggressive methods are preferred for studying the organic matrix of dental tissues in proteomics. Saline, ethanol, sodium hypochlorite, or thymol solutions are commonly used for storage, whereas freezing is typically preferred when maximum preservation of the organic phase is required (31–33).

Freezing at −80°C is typically considered the gold standard for storing biospecimens, as both enzymatic and non-enzymatic degradation of proteins are significantly reduced at this temperature (34). However, studies have shown that biomolecules can still undergo alterations at −80° C, potentially affecting the organic components, including proteins and nucleic acids, when stored for extended periods (35–37). In our study, the protein composition was comparable between the −80°C and EtOH groups, with similar overall protein composition and relative abundance of most relevant proteins, based on PantherDB protein classes. Therefore, storage in 70% ethanol at room temperature offers a practical and effective alternative to −80°C, especially in clinical settings where immediate freezing is often not feasible. Ethanol storage could also reduce the shipment costs in multicenter studies and provide favorable antibacterial properties.

Our results demonstrate better conservation of the proteome in the EtOH and −80°C groups compared to the PBS and Air groups. Although the −80°C provided the most diverse proteome when compared to all other conditions, there are notable differences in the preservation of specific proteins. For instance, the relative quantities of enamel-specific proteins such as MMP20 and ameloblastin were lowest in the −80°C group compared to all three conditions (Figure 2B). The relative abundances of collagen and KLK4 in air-dried samples were comparable to those stored at −80°C, and amelogenin levels were even higher. Similarly, air-dried samples showed an increase in the overall percentage of the extracellular matrix protein category (PC00102), while the other categories were similar to those in the −80°C group (Figure 3). These findings suggest that extracellular matrix proteins are more resilient to protein degradation due to environmental changes; however, it is unclear whether and to what extent residual enzymes contribute to this degradation.

Saline solutions, such as Hank's Balanced Salt Solution and PBS, are commonly used in both clinical and laboratory settings to preserve cell viability on tooth surfaces, including periodontal ligament cells. However, in this study, the PBS and Air groups had the lowest total number of proteins, which was less than half the number detected in the −80°C and EtOH groups. The temperature of the solutions in our study (RT) could have affected the results, and storing PBS at 4°C rather than RT might increase the yield. Interestingly, relative abundances of key proteins were lowest in PBS, even when compared to Air.

We observed a high abundance of proteins that are not specific to enamel (e.g., from blood and surrounding tissues) in all groups, as might be expected in unerupted, extracted teeth. These non-enamel-specific proteins have been previously reported (38–40). The high level of these proteins in the EtOH and Air groups in this study suggests either their deeper penetration into enamel or the preservation of non-enamel-specific proteins under these conditions.

In the case of proteins considered exogenous, such as serum albumin, the storage medium becomes especially important when these proteins are studied and their role is considered in the pathophysiology of developmental dental defects (15).

A limitation of our study is the short-term storage of all samples at −80°C for transport. Although all samples were subjected to the same step, hereby setting a baseline, the freezing followed by another storage condition might result in a difference in preservation compared to the absence of freezing. Another limitation is that we did not test the variability of protein content between the four quarters of the unerupted teeth, assuming that protein composition would be sufficiently similar in all four quarters of a single tooth crown, based on our understanding of the processes and patterns of tooth crown formation (41, 42). Further investigation looking into the effect of storage conditions could also focus on the post-translational modifications of the enamel proteome.

Our findings can inform future studies on selecting the optimal sample storage medium, taking into account both logistical considerations and specific research questions. Studies focusing on the albumin content in hypomineralized enamel may require a different storage method than those interested in MMP20 levels in developing enamel. Different storage conditions not only significantly alter the abundance of specific protein groups but also make it difficult to compare data between studies. For example, the difference in protein content in enamel between studies can be attributed to the varying storage conditions for the analyzed teeth (25, 40).