In humans there are 11 genes annotated as TMED1-11. TMED family members are expressed throughout the body, as demonstrated in mice (Strating et al., 2009), and are highly expressed in secretory cell types (Zhang and Volchuk, 2010). Given their ubiquity, it is thus unsurprising that some TMED proteins are developmentally essential and knockout of either TMED2 or TMED10 is embryonic lethal in mice (Denzel et al., 2000; Jerome-Majewska et al., 2010) and reduces viability in cultured cells (Blomen et al., 2015).

Various studies have shown that the TMED proteins form oligomers of varying stoichiometry (Fullekrug et al., 1999; Emery et al., 2000; Jenne et al., 2002). A series of siRNA knockdown experiments revealed that knockdown of TMEDs 2, 4, 5, 9, or 10 destabilized other TMED family members while TMED7 knockdown primarily affected TMED5. Loss of TMEDs nine or 10 inhibited GPI-AP trafficking, whereas WNT trafficking was inhibited in cells lacking either TMEDs 2, 4, 9, or 10 (Tashima et al., 2022). Because of the interdependency between TMED family member expression and function, it is technically challenging to discriminate between the functions of individual TMED proteins or their oligomeric complexes.

Mammalian TMED9 and its yeast homolog, Erv25p, were first discovered as secretory pathway proteins (Dominguez et al., 1998; Marzioch et al., 1999). TMED9 localizes primarily to the ER and ERGIC, but is found in post-Golgi secretory vesicles along with other TMED family proteins (Shevchenko et al., 1997; Dominguez et al., 1998; Marzioch et al., 1999; Breuza et al., 2004). Later TMED9 was discovered to be critical for the generation of ER exit sites (ERES) in a cell-free microsome budding assay (Lavoie et al., 1999). Further emphasizing its role(s) in the secretory pathway, depletion of TMED9 leads to the fragmentation of Golgi structures and the partial dissociation of COPI from the Golgi (Mitrovic et al., 2008). The yeast homolog of TMED9, Erv25p, has been shown to be play a role in efficient ER-to-Golgi transport of the yeast GPI-AP, Gas1 (Belden and Barlowe, 1996). However teasing apart TMED9’s individual role from other TMED-family members, including TMEDs 2 and 10, is difficult because knockdown of each impacts the expression of the others (Fujita et al., 2011). Taken together, TMED9 along with its family appears to regulate multiple critical trafficking steps in the secretory pathway. Excellent reviews have been written to discuss the role of the TMED proteins in the early secretory pathway (Pastor-Cantizano et al., 2016; Aber et al., 2019).

Recently, TMED9 was shown to participate in unconventional protein secretion (UPS) from the ER to the plasma membrane during ER stress in cells expressing the dominant-inhibitory form of ADP-ribosylation factor 1 (ARF1-Q71L), which blocks ER-to-Golgi transport (Park et al., 2022). TMED9 was found to participate in the assembly of a heterooligomeric trafficking complex governing SARS-Cov2 spike protein and cystic fibrosis transmembrane conductance regulator (CFTR) secretion (Park et al., 2022). Although Park et al. (2022) found that TMED9 did not bind to CFTR or Spike proteins, silencing TMED9 reduced the cell surface trafficking of these UPS cargo. These findings suggest that TMED9 may participate in a variety of yet undiscovered trafficking pathways.

The TMED proteins are structurally conserved among eukaryotes despite significant variations in sequence identity (Strating et al., 2009) (Figure 1A). Each family member contains four major regions: the GOLD domain, coiled-coil domain, transmembrane domain, and a cytoplasmic COP-binding region (Figures 1B, C). Whether these conserved domains allow the TMED proteins to act interchangeably in certain processes is unknown.

Elevated TMED9 expression has been observed in multiple cancer types (Ju et al., 2021). In breast cancer, elevated TMED9 levels are associated with poor prognoses (Ju et al., 2021). In head and neck squamous cell carcinoma, expression of each of the TMED proteins is elevated. In particular, high expression of TMEDs 2, 9, and 10 was found to be associated with poor prognoses, whereas high expression of TMEDs 1, 3, 4, 5, 6, and 7 was not (Gao et al., 2022). Similarly, elevated TMED9 expression is associated with reduced survival time in individuals with epithelial ovarian cancer (EOC) in vivo. TMED9 knockdown reduces EOC cell proliferation in vitro (Han et al., 2022).

TMED9 expression may regulate cancer cell proliferation through its effect on growth factor signaling (Buechling et al., 2011; Nakano et al., 2017; Zhang X. et al., 2020; Di Minin et al., 2022; Tashima et al., 2022). For example, biochemical and microscopy approaches revealed that TMED9 loss was associated with dysregulation of TGFɑ trafficking and secretion in colon cancer and hepatocellular carcinoma cells (Mishra et al., 2019; Yang et al., 2021). Furthermore, loss of TMED9 led to impaired WNT trafficking (Tashima et al., 2022) and significant changes in the expression of genes regulated by WNT signaling (Yang et al., 2021). This TMED9-WNT signaling axis has been implicated in Paneth cell function in the intestines (Goga et al., 2021).

The TMED proteins have been implicated in various neurodegenerative diseases. TMED10 associates with and is required for the clearance of artificial and prion-disease associated mutants of prion protein (PrP) (Satpute-Krishnan et al., 2014). TMED2 was found to co-immunoprecipitate with atlastin whose misfolding leads to hereditary spastic paraplegia (Namekawa et al., 2007). Although TMED9 has not been thoroughly studied in the context of neurological disease, TMED9 has been shown to interact with wild type TDP-43, whose aggregation has been associated with the development of amyotrophic lateral sclerosis (ALS) (Redler and Dokholyan, 2012; Feneberg et al., 2020), and to associate with spastin whose mutations lead to hereditary spastic paraplegia (Reid et al., 2005). The precise motifs or domains of TMED9 involved in binding to TDP-43 or spastin remain unknown.

Various studies have demonstrated that alterations in TMED9 and TMED10 (commonly referred to as TMP21 in the Alzheimer’s field) expression promote the processing of amyloid precursor protein (APP) to amyloid beta (Aβ) by γ-secretase. Alzheimer’s disease has been associated with mutations in the genes encoding subunits of γ-secretase (Zhang et al., 2011). A single nucleotide polymorphism in TMED10 that resulted in heightened TMP21 expression was found to be genetically associated with Alzheimer’s disease in patients (Zhang et al., 2018). Additionally, alterations in TMED10 expression were found to impact pathological APP processing in cell culture models (Zhang et al., 2018). TMED10/TMP21 has been shown to co-immunoprecipitate with and regulate the activity of the γ-secretase complex. Intriguingly, depletion of TMED10/TMP21, results in increased generation of Aβ (Chen et al., 2006; Dolcini et al., 2008). Similarly, TMED9 co-immunoprecipitates with the core γ-secretase components and knockdown of TMED9 mRNA induces an increase in Aβ generation (Hasegawa et al., 2010; Bai et al., 2015). Because of TMED9’s tendency to heterooligomerize with TMED10/TMP21 (Dominguez et al., 1998; Fullekrug et al., 1999), TMED9 may function in a complex with TMED10/TMP21 to regulate γ-secretase processing of APP.

The proteinopathy mucin-1 kidney disease (MKD) results from a frameshift mutation in the MUC1 gene. (Dvela-Levitt et al. (2019) recently demonstrated that TMED9 binds to and mediates the post-ER trafficking of MUC1 aggregates. Under steady-state conditions, this TMED9-MUC1 complex drives the accumulation of toxic MUC1 in the ERGIC. Fortuitously, the authors found that the small molecule BRD4780 was able to reduce MUC1 aggregate levels both in vivo in mice and in vitro in cell culture and human kidney organoid models by reducing TMED9 stability and accelerating clearance of TMED9-MUC1 complexes from the ER and ERGIC to lysosomes (Dvela-Levitt et al., 2019).

The TMED protein family field is a rapidly evolving area of research. The exciting discovery that BRD4780 targets pathological TMED9-MUC1 aggregates to lysosomes demonstrates the potential to pharmacologically target TMED proteins for the resolution of proteinopathies (Dvela-Levitt et al., 2019). It is as yet unclear how BRD4780 induces lysosomal degradation of TMED9-MUC1, but possible mechanisms may involve directly altering TMED9’s structure or by preventing its oligomerization with other TMED family members. Preventing hetero-oligomerization of TMED9 has been shown to reduce its stability and the stability of other TMED proteins (Tashima et al., 2022). Because of this interdependence, BRD4780 may potentially be exploited to modulate various PQC pathways involving TMED heterooligomers. These findings encourage future research into the role of TMED9 and the other TMED proteins in clinical proteinopathies.

Beyond their clinical implications, the TMED proteins have now been shown to participate in a variety of essential cellular processes. New empirical tools such as cryoelectron microscopy and AI based modeling tools like Alphafold2 enable future structural studies to better characterize the interaction of the TMED proteins with one another and their cargo. Furthermore, structure-driven mutagenesis strategies may be used to disrupt oligomer formation in order to reveal the independent functions of the TMED proteins in the secretory pathway. Compelling questions include: Does environmental pH affect the function and binding of TMED proteins? Does N-linked glycosylation influence TMED cargo recognition? What is the precise role of TMED9 in autophagosome biogenesis, and how is this balanced with its role in secretion? What are the specific roles of each TMED family member and their various oligomeric states? As we gain answers, we may soon understand the functions and clinical relevance of TMED9 and its stubbornly mysterious family of proteins.