Biosynthesis, processing, cellular localization and function of the mutants LPH-Y1473X and LPH-D1796fs
Until present seven mutations in the coding region of LPH have been identified in Finnish families. Recent genetic testing has revealed two novel mutations, Y1473X and D1796fs, which were found as compound heterozygous pattern in a Japanese infant with CLD [3]. The genotype/phenotype relationship is examined in this paper at the biochemical and cell biological levels. Both mutations are located in domain IV of the extracellular region of LPH compatible with partial truncation of homologous domain IV and complete elimination of the entire membrane anchor as well as the cytoplasmic tail.
Since both mutations appeared in a compound heterozygous mode in one patient, we were interested in investigating the influence of each single mutation per se on the enzymatic function and intracellular transport events of LPH.
For this purpose the mutations, Y1473X and D1796fs, were introduced into the coding region of wild type LPH separately (thereafter referred to as LPH-Y1473X and LPH-D1796fs) and the generated mutants were expressed in COS-1 cells. Detergent extracts of the biosynthetically-labeled transfected cells were immunoprecipitated. To assess the differences in maturation state and glycosylation pattern among wild type LPH, LPH-Y1473X and LPH-D1796fs, the immunoprecipitates were treated with endoglycosidase H [21]. Endo H cleaves exclusively mannose-rich and some hybrid types of N-glycans which exist in the ER up to cis-Golgi. Figure 1A shows that wild type LPH revealed a 215 kDa mannose-rich glycosylated, endo H-sensitive band and a 230 kDa endo H-resistant complex glycosylated band. By contrast, LPH-Y1473X and LPH-D1796fs revealed exclusively endo H-sensitive mannose-rich protein bands respectively smaller than their untreated forms. We further addressed the trafficking of the two mutant proteins by performing pulse chase experiments. For this purpose, COS-1 cells were biosynthetically labeled for 2 h and chased for 0, 4, 8 and 12 h. Figure 1B shows that wild type LPH reached the status where 50% of the protein is mannose-rich glycosylated and 50% is complex glycosylated after 4 h of chase. After 12 h of chase, only the mature form was detectable. The mutated protein variants, LPH-Y1473X and LPH-D1796fs, persisted as mannose-rich glycosylated proteins, but remarkably the bands disappeared almost completely after 8 h of chase. To examine the possibility whether the mutants are ultimately secreted, the media of the biosynthetically labeled COS-1 cells were collected and immunoprecipitated. However, neither the wild type LPH nor the two mutated proteins exhibited any secreted forms (data not shown). These results indicate clearly that the mutants persist in their mannose-rich glycosylated forms, likely in the ER, and are ultimately degraded.
Persistence of the two mutants as mannose-rich glycosylated species is suggestive of an ER localization of the two protein variants. To substantiate these results using a different procedure, we analyzed the intracellular localization of the mutants by confocal immunofluorescence microscopy. As shown in Figure 2, wild type LPH is located in the ER and the Golgi apparatus as well as at the cell surface, while both mutant variants were predominantly located in the ER as assessed by the typical ER net-like structures.
The lactase active site is located at Glu1749, while that of phlorizin-hydrolase at Glu1273 [12]. The mutant LPH-Y1473X lacks the lactase activity site and it was therefore expected that this mutant reveals no lactase activity, which we confirmed. The mutant LPH-D1796fs contains the lactase active site, but its activity towards lactose was undetectable (data not shown). These results indicate that the compound heterozygote pattern of these two functional-inactive mutants is responsible for complete lactose intolerance in the infant.
Potential effects of the mutants LPH-Y1473X and LPH-D1796fs on wild type LPH in a heterozygote background
CLD in the patient was initially suggested by typical malabsorption symptoms upon milk uptake and subsequently through the identification of the mutations Y1473X and D1796fs on each allele of the LPH gene.
The compound heterozygote pattern of inheritance in this CLD case raises questions related to potential lactose tolerance in the parents, who are heterozygote carriers of one of the defective LPH alleles. We addressed therefore the question whether a single normal parental allele in conjunction with the diseased one is sufficient to produce lactase protein that has adequate digestive capacity towards dietary lactose. The rationale for this assumption is the dimeric quaternary structure of wild type LPH that is generated in the ER and warrants enzymatic activity as well as transport competence of LPH [16]. Provided that minimal folding requirements are fulfilled, such as correct folding of domains involved in dimerization of LPH, it should be still considered that two different protein isoforms, such as a wild type LPH and a truncated LPH mutant can form heterodimers, which may regulate the enzymatic function. We mimicked therefore the in vivo situation by co-expression of each mutant separately with the wild type LPH, assessed their potential interaction in co-immunoprecipitation experiments and analyzed the activities of the immunoprecipitated LPH. Figure 3 shows protein bands corresponding to wild type LPH that were obtained by immunoprecipitation of cellular lysates from co-transfected and biosynthetically-labelled cells. Since neither the LPH-Y1473X nor LPH-D1796fs mutants appeared in the same electrophoretic lane we conclude that these mutant forms did not co-immunoprecipitate or interact with the wild type LPH.
The enzymatic activities of LPH in the co-transfected experiments did not change or support the notion that interaction has taken place. This in turn clearly indicated that wild type LPH retained its full activity since it has generated its enzymatically active homodimers.
Mutant LPH-Y1473X or LPH-D1796fs are not temperature-sensitive
Several mutations in proteins that are associated with protein folding diseases exhibit a temperature sensitive pattern, in which lowering the temperature leads to a recovery of correct folding and further protein trafficking out of the ER [5,23-25]. To determine if the mutants are temperature-sensitive and can exit the ER to the Golgi apparatus in a time-delayed manner, a pulse chase experiment was performed at 20°C. The control utilized a similar protocol except that the temperature was changed to 37°C. The wild type LPH was detectable as a mannose-rich glycosylated and a complex glycosylated protein band after 6 h and 18 h of chase at 20°C incubation temperature (Figure 4). At this temperature proteins are predominantly blocked in the Golgi apparatus [26]. This becomes evident in the 18 h chase time point, which shows substantial increase in the intensity of the complex glycosylated mature protein that is processed in the Golgi apparatus. In contrast to the wild type LPH, both mutant proteins appeared as mannose-rich glycosylated species after 6 h of chase at 20°C and did not reach a further maturation status after 18 h of chase, since no complex glycosylated band appeared. The same experiment was performed at 37°C. Interestingly, wild type LPH was still detectable after 18 h of chase, mainly as the mature complex glycosylated form, but LPH-Y1473X or LPH-D1796fs disappeared completely suggesting that the mutants are subject to a degradation mechanism in the ER.