Clinical and neuropathological similarities in NCL disorders may result from functional redundancy or co-operation of different NCL proteins. Initial evidence of co-operation has previously been obtained from activity measurements and gene expression analyses. For example, CLN2/TPP1 activity has been shown to be elevated in other forms of NCL [23, 36, 37]. Furthermore, the mRNA expression levels of a variety of NCL genes have been reported to be altered in different NCLs . Recent studies of animal models have further supported a common mechanism in the disease pathogenesis of the NCLs. Selective loss of interneurons, early defects on thalamocortical neuron survival as well as early glial responses are detected in different mouse models as well as in large animal models [8, 38]. Molecular interactions between NCL proteins have also been reported previously [23, 39]. Here we continued the search for NCL protein interactions and their intracellular consequences. We demonstrate that CLN5 has molecular connections to at least to five other NCL proteins, namely CLN1/PPT1, CLN2/TPP1, CLN3, CLN6 and CLN8, suggesting a central role for CLN5 in the NCL network.
Mutations of CLN5 and consequent trafficking defects can result in functional consequences on the interacting proteins, as well as in changes in their distribution. Most NCL proteins are not, however, detectable endogenously by immunofluorescence analyses using the currently available antibodies and therefore this hypothesis could not be tested in patient fibroblasts. However, simultaneous overexpression of binding partners with the modified, transport incompetent CLN5-TD (CLN5-flag330) construct suggested that the interaction between PPT1/CLN1 and CLN5 is strong and occurs already in the ER, where the interactions between CLN5 and the two ER resident NCL proteins, CLN6 and CLN8, would also naturally occur. Since CLN5 resides in the lumen of intracellular organelles, interaction between CLN5 and the transmembrane proteins must be mediated by the lumenal domains of these proteins. The finding that NCL interactions can occur already in the ER is important since previous studies have also suggested that the ER is an important organelle for the function and/or trafficking of NCL proteins. For example, CLN3 carrying the most common JNCL mutation  has been reported to preserve a significant function in the ER. Although the mutated CLN3 is retained in the ER, it was shown to be able to affect the size of lysosomes . Furthermore, our data suggest that TPP1/CLN2, another lysosomal enzyme, is likely to interact with CLN5 only in the late endosomes/lysosomes. This was supported both by the interaction analysis and the unaffected transport of CLN2. Therefore, the present data suggest that interactions between CLN5 and other NCL proteins can occur along the secretory pathway, and the interactions are not strictly dependent on the steady-state localization or the solubility of the NCL proteins.
Based on our trafficking experiments, the main focus was on the interaction between PPT1/CLN1 and CLN5. Unlike the transport deficient, strictly ER-resident CLN5-TD, the ER exit competent CLN5FinMajor did not prevent the lysosomal trafficking of PPT1 but rather, PPT1 was able to facilitate the trafficking of the mutated CLN5 from the ER and Golgi to the lysosomes. Lysosomal proteins have also previously been shown to assist each other in their lysosomal trafficking. For example, the CLC7 chloride transporter, involved in an NCL-like disorder in mouse, has been shown to facilitate the transport of Ostm1 to the lysosomes, where the proteins act together in lysosomal chloride transport . It was also recently demonstrated that the lysosomal membrane protein LIMP-2 is required for the mannose 6-phosphate receptor-independent targeting of β-glucocerebrosidase. Interestingly, the overexpression of LIMP-2 was also able to facilitate the transport of the mutated, trafficking deficient β-glucocerebrosidase from the ER to the lysosomes . Both PPT1 and CLN5 are soluble intravesicular proteins and are not able to interact with cytoplasmic sorting and transport machinery per se, like CLC7 and LIMP-2. However, trafficking of CLN1/PPT1 has been reported to show properties different from classic lysosomal enzymes  and we have recently discovered that also CLN5 can use M6PR-independent pathways for its lysosomal trafficking (our unpublished observations). Therefore, it is possible that the formation of the CLN1-CLN5 complex may be important for utilizing other trafficking pathways than the classical M6PR pathway. In which circumstances this is required in vivo, remains to be studied in further experiments.
A close connection between PPT1/CLN1 and CLN5 has already been suggested in previous studies. The proteins share similar expression patterns in the mouse brain and in the prenatal human brain [16, 21, 44, 45]. Our recent global gene expression profiling analyses of the Cln1-/- and Cln5-/- mouse brains implicated a common defective pathway mediated by phosphorylation and potentially affecting the maturation of axons and neuronal growth cones . Here, we provide further evidence for a tight relationship between the two proteins by showing not only the interaction between the proteins but also, demonstrating significantly increased expression levels of Cln1 mRNA in the Cln5-/- mouse brain tissue. This suggests a possible compensatory role for PPT1 in CLN5 deficiency. Functional connection of CLN5 and PPT1 is also suggested by the shared interaction partner not belonging to the NCL protein family, the F1-ATP synthase. The ectopic F1-ATP synthase has been shown to function as an apoA-I receptor on the plasma membrane  and both the amount of the F1-complex as well as the uptake of apoA-I have been shown to be increased in Cln1-/- mouse neurons . Therefore, both PPT1 and CLN5 could be connected to the maintenance of lipid homeostasis. In general, accumulating evidence has indicated dysregulated lipid metabolism in different forms of NCLs and several NCL proteins have been functionally linked to lipid metabolism [[48, 49], reviewed in [50–52]].