Diagnosing mucopolysaccharidosis IVA

Adequate sample quality is essential for urinary screening. Dilute urine samples can be particularly problematic and may not be accepted by the laboratory. A first morning void or 24 h collection is preferred and a minimum of 15 mL of sample is usually recommended; however, patient age must be taken into consideration. As shipping, handling, and logistical considerations for urine samples vary, clinicians should contact the laboratory for specific details.

Urinary screening is based on the excretion of excess GAGs that occurs in the MPS disorders. The urinary excretion of GAGs, when normalized to creatinine, is high in infants and young children, decreases with age, plateaus in the second decade of life, and remains constant through adulthood (de Jong et al 1989; Whitley et al 1989a; Piraud et al 1993; Gray et al 2007; Wood et al 2012). Evaluations can be done for both the total amount of GAG excreted (quantitative analysis) and the specific types of GAG excreted (qualitative analysis). In the case of MPS IVA, evaluating both simultaneously is critical to avoid false negative results as KS can be present without elevating the total amount of GAG above the upper limits for unaffected individuals (Whitley et al 1989a; Piraud et al 1993; Tomatsu et al 2004; Gray et al 2007). Determining both total urinary GAG quantity and GAG identity is strongly recommended, especially if MPS IVA is considered a possibility.

Some of the methods for analyzing total urinary GAG used in the past, such as spot tests and turbidity tests, are no longer recommended. Spot tests for MPS have very poor specificity (12–29 % false positives) and sensitivity (19–35 % false negatives) and do not allow for interpretation of results by patient age or sample concentration (Pennock 1976; de Jong et al 1991). The cetylpyridinium chloride (CPC) assay, a turbidity test based on GAG precipitation, also has a high false negative rate reported to range from 11 to 30 % depending upon the pH, with MPS IVA being the most commonly missed MPS (de Jong et al 1989). Another test for total urinary GAG, the uronic acid-carbazole test is particularly unsuited for use when MPS IVA or B is suspected because it is based on the measurement of uronic acid in GAGs; however, KS, the main GAG excreted in MPS IVA and B, does not contain uronic acid residues. Therefore, MPS IVA and B cannot be reliably detected by the uronic acid-carbazole test (de Jong et al 1989; Frazier et al 2008).

Currently, the analysis of total urinary GAG is usually performed in a quantitative manner by spectrophotometric analysis of urine combined with a cationic blue dye such as dimethylmethylene blue (DMB) (Whitley et al 1989b). Total GAG is reported relative to creatinine to normalize for urinary concentration and age-based reference ranges are used to interpret results (de Jong et al 1989). As previously discussed, although most MPS IVA patients do excrete KS, many do not excrete enough to elevate their total urinary GAG level above the reference range for unaffected individuals. This results in an unacceptably high false negative rate when this quantitative analysis is performed alone, 15 % for MPS in general and higher for MPS IVA in particular (Whitley et al 1989a; Tomatsu et al 2004; Gray et al 2007). To reduce false negative rates, running a qualitative analysis simultaneously is strongly recommended (Piraud et al 1993; Gray et al 2007).

Qualitative analyses isolate GAGs from urine then separate them using either thin layer chromatography or electrophoresis (Wessler 1968; Humbel and Chamoles 1972). GAG identity is determined by relative position on the plate or gel. While the results are not quantitative, they can provide substantial insight into which MPS disorder is present and may reveal the presence of KS for some patients whose total urinary GAG levels are not elevated (Piraud et al 1993). Unfortunately, KS can be challenging to separate and visualize. Separation can be improved by using either multiple-step 1D electrophoresis (Hopwood and Harrison 1982) (Fig. 3) or 2D electrophoresis (Hata and Nagai 1972) (Fig. 4). However, despite the improved separation, interpretation of results is still subjective and can be challenging (Fig. 5).

Tandem mass spectrometry detection (MS-MS) and quantification of several GAG species or KS alone is an attractive alternative to the methods previously described (Oguma et al 2007; Ohashi et al 2009; Tomatsu et al 2010; Hintze et al 2011). A new method for quantifying urinary KS alone, as opposed to total urinary GAG, has recently been described (Martell et al 2011) and is based on the method that has been previously published (Oguma et al 2007). Keratanase II is used to digest KS and the resulting disaccharides are analyzed by LC-MS/MS. While currently used (non MS-MS) methods of urinary GAG analysis are either qualitatively specific for KS but non-quantitative or non-specific for KS and only quantitative for total GAG, this new method has the advantage of being both KS-specific and quantitative. Additionally, the LC-MS/MS method is likely to provide higher sensitivity than other methods. Reported assay validation results show separation between MPS IVA patient (n = 160) and unaffected control (n = 186) KS ranges in urine and provide compelling evidence for the clinical utility of this method. However, the degree of separation is small, particularly for older patients (0.2–1.3 (unaffected) vs. 1.7–37.9 (affected) μg/mg creatinine for patients >15 years of age) (Martell et al 2011). The increased sensitivity and specificity of this technology has provided renewed interest in the analysis of KS in blood. While some research suggests that blood KS may be an additional biomarker of interest in MPSIVA patients, further research is needed to understand its use and determine the best methods for analysis (Oguma et al 2007; Martell et al 2011; Hintze et al 2011). Finally, another new method has been described for analysis of non-reducing ends of urinary GAGs by LC-MS. Currently this method does not provide a measure of non-reducing ends for MPS IVA (Lawrence et al 2012). MS/MS methodology for the analysis of KS is not widely available in diagnostic, clinical settings which limit its current usefulness to physicians. However, with increased interest in MS-MS assays and the wider availability of the required reagents, MS-MS methodologies are likely to have a large impact in the diagnosis of and monitoring of MPS IVA in the future.

Ultimately, while both quantitative and qualitative urinary GAG analyses can provide insight into the diagnosis, they are still considered screening methods and can produce false positive and false negative results (Piraud et al 1993). The relatively high rate of false negatives is particularly problematic in mild cases where a clinician may not suspect MPS IVA and is interested in using urinary analysis to rule out the diagnosis. It is this testing strategy in patients with borderline or normal total urinary GAGs and/or KS that often leads to the diagnosis being missed or delayed. While most MPS IVA patients excrete KS (Tomatsu et al 2004; Martell et al 2011), the absence of KS excretion in MPS IVA has been reported in the past (Fujimoto and Horwitz 1983). This may have been due to the limited sensitivity of the technology available at the time. However, regardless of the technology used, urine analysis is likely to be more challenging in older patients especially those with the mild form of the disease who do not present with symptoms until adolescence. One of the main sources of KS is cartilage, so as bone growth centers close in patients reaching puberty, the rate of KS accumulation slows and urinary excretion is substantially reduced and may even disappear (Longdon and Pennock 1979). Therefore, false negative results become more likely as patients age. Increased disease awareness may also lead to increased false negative rates as milder cases with lower KS levels are screened more often. Whether or not the new LC-MS/MS method will be able to overcome these challenges remains to be seen. Regarding the specificity of urinary analysis for MPS IVA, patients with MPS IVB and multiple sulfatase deficiency (OMIM # 272200) also excrete KS (Arbisser et al 1977; Groebe et al 1980) and, although unexpected, elevated KS has been reported in other MPS disorders and mucolipidoses (Tomatsu et al 2005b). KS can also be elevated in other lysosomal disorders such as Fucosidosis (OMIM # 230000) and GM1 gangliosidosis (OMIM #230500, 230600, 230650) which is caused by the same enzyme deficiency as MPS IVB (Ullrich and Kresse 1996). Hence the absence of KS does not conclusively rule out MPS IVA and the presence of KS does not conclusively diagnose MPS IVA. Proceeding to enzyme activity analysis even if urinary analysis results are negative is strongly recommended if there is clinical suspicion of MPS IVA.

In addition to being used to screen for MPS, urinary analysis is also commonly used in other MPS disorders to track a patient’s response to enzyme replacement therapy (ERT) (Wraith et al 2004; Harmatz et al 2006; Muenzer et al 2006). Should ERT become available for MPS IVA, using urinary analysis to track MPS IVA patient responses may present a significant challenge. New methodology, such as the LC-MS/MS assay, or possibly even new biomarkers, may be required to successfully quantify biochemical patient outcomes.

Diagnosis of MPS IVA requires the demonstration of a deficiency in GALNS activity. Fibroblasts and leukocytes are the recommended sample types for this analysis. Prenatal samples, such as dissected chorionic villi, cultured chorionic villus cells, and amniocytes, can also be used, facilitating prenatal diagnosis (Zhao et al 1990; Kleijer et al 2000). Additionally, protocols for evaluating GALNS activity in a DBS have recently been proposed as screening methods (Duffey et al 2010; Camelier et al 2011).

Fibroblast samples are recommended for enzyme activity analysis because the impact of environmental and logistical factors during shipment can be minimized and corrected through culturing of the cells in the receiving laboratory. The culturing process also allows for the generation of a large number of cells for analysis and more cells can be grown at a later time if further analyses are needed. On the other hand, skin punch biopsies are invasive and a significant amount of time is required prior to analysis to culture the cells (ranging from 2 to 10 weeks) which can occasionally fail to grow or become contaminated. The total time from sample collection to the reporting of results to the clinician can be relatively lengthy, particularly if problems are encountered during cell culture and the culturing process has to be repeated. Some laboratories also require that the fibroblasts from the skin punch biopsy are cultured prior to being sent to the laboratory which requires considerable time, effort, and expertise on the part of the establishment sending the sample.

Leukocytes isolated from whole blood allow for more rapid analysis as cell culture is not required. Typically, the time from sample collection to the reporting of results to the clinician is less than 2 weeks. Whole blood samples normally provide a sufficient number of cells for analysis; however, additional cells cannot be generated by cell culture, so another sample must be obtained if the initial sample is insufficient or if the results are inconclusive. Furthermore, leukocyte samples are susceptible to environmental extremes during shipping in countries with a hot climate (Burin et al 2000). Sample deterioration may also become an issue when shipping long distances results in a prolonged amount of time spent in transit. This has lead many laboratories to require arrival of the sample within 24 to 48 h post-draw. Ultimately, the measurement of lysosomal enzymes in blood requires careful attention to the quality of the incoming sample including measurement of controls to verify sample integrity.

DBS samples are a convenient sample type both in regions where it is difficult to ship whole blood or tissue samples and for newborn screening programs. However, as described for leukocytes, proper sample collection and shipping are critical to the success of the analysis (Camelier et al 2011). A video on DBS collection technique is available online (Fundación para el Estudio de las Enfermedades Neurometabólicas 2011). The date of collection should always be noted on the card to aid in interpretation of results. The card should be dried at room temperature for at least 4 h. Due to the temperature sensitive nature of GALNS in a DBS (Camelier et al 2011), cards should be stored at 4°C after drying and shipped promptly; the longer the period of time between collection and analysis, the higher the risk of a false positive result. If high temperatures are possible during shipment, the card should be shipped in an insulated container with ice packs (Camelier et al 2011).

While measurement of GALNS activity in DBS samples is very useful for screening, this method is not as robust as it is in fibroblasts or leukocytes due to the lower number of cells present in the sample. Furthermore, more data are needed to evaluate GALNS stability in a DBS, especially since DBS samples are more likely to be exposed to environmental extremes during shipping than leukocytes. For these reasons, DBS results should not be used alone to reach a diagnosis. Confirming positive DBS results by enzyme activity analysis in fibroblasts or leukocytes is strongly recommended. Alternatives are discussed in the section entitled “Reaching a diagnosis”.

It is important to note that if a patient has recently received a blood transfusion or a bone marrow/stem cell transplant, enzyme activity analysis in both leukocyte and DBS samples could give inaccurate results. If there is any concern for this type of interference, consultation with the laboratory is recommended or, at minimum, this information should be included on the sample requisition. Fibroblast samples will not be impacted by these treatments and are recommended in these instances.

GALNS acts on two substrates, N-acetylgalactosamine-6-sulfate (GalNAc-6S) (Matalon et al 1974) and galactose-6-sulfate (Gal-6S) (Glossl and Kresse 1982; Yutaka et al 1982). GalNAc-6S is a component of C6S and Gal-6S is a component of KS. Both of these substrates are used currently to demonstrate deficient GALNS activity. The GalNAc-6S based assay uses radio-labeled natural substrate (Glossl and Kresse 1978) and the Gal-6S based assay uses fluorogenic artificial substrate (van Diggelen et al 1990). The radio-labeled substrate, a tritiated disulfated trisaccharide prepared from C6S, was developed first (Glossl and Kresse 1978). Using this substrate, GALNS activity is determined based on the amount of radioactivity released from the substrate with a lack of GALNS activity resulting in low generation of signal. Although this method is still in use, many laboratories have opted to switch to an assay based on a fluorogenic substrate (van Diggelen et al 1990). The fluorogenic assay uses 4-methylumbelliferyl-β-D-galactopyranoside-6-sulfate (4MU-Gal6S) as the substrate and is accomplished in two steps. First, GALNS present in the sample removes the 6-sulfate, and then exogenous β-galactosidase removes the galactoside, freeing the 4-methylumbelliferone adduct, which fluoresces under high pH. The addition of exogenous β-galactosidase to the reaction mixture is critical because conditions in which β-galactosidase is deficient (MPS IVB, I-cell disease (OMIM #252500), and GM1 gangliosidosis) would result in significant GALNS activity underestimation and possibly misdiagnoses (van Diggelen et al 1990). Removal of endogenous phosphates and sulfates is also important as they are lysosomal sulfatase inhibitors (van Diggelen et al 1990). Modification of the original published protocol (van Diggelen et al 1990) by increasing the substrate concentration ten-fold is recommended to increase assay sensitivity.

In addition to these two currently used methods, a tandem mass spectrometry based method using a novel substrate has recently been developed (Duffey et al 2010). Incorporation of this new assay into a multiplexed lysosomal storage disease panel for use in newborn screening programs is being considered (Khaliq et al 2011).

Regardless of the methods used, performing enzyme activity analysis for MPS IVA should also involve the evaluation of additional enzymes to control for sample integrity and to rule out differential diagnoses.

To confirm that low GALNS activity is not due to sample degradation, the activity of a reference enzyme with similar stability in the same sample should be measured. If a DBS is used, measuring a reference enzyme in the same sample to confirm integrity is still recommended, but may not be sufficient to rule out an effect of handling on GALNS activity because the stability of GALNS as compared to other enzymes in a DBS is not known. Simultaneously collecting and shipping a negative control DBS sample (from an unaffected healthy individual) to confirm sample handling has not compromised GALNS activity is recommended (Camelier et al 2011). However, obtaining control samples may be difficult logistically for some laboratories, particularly for routine testing of a large number of patients. Parental or familial samples are not recommended for this purpose as their GALNS activity level may be below normal if they are carriers.

It is also important to confirm that the low GALNS activity is not being caused by a different disease. Multiple sulfatase deficiency (MSD) and mucolipidosis II or III (ML II/III) also impact GALNS activity. MSD is a disease in which the activities of several sulfatases, including GALNS, are deficient (Dierks et al 2003) and ML II/III impair mannose-6-phosphate guided enzyme targeting causing lysosomal enzyme activities to be low in fibroblasts, elevated in plasma, and relatively unaltered in leukocytes (Neufeld and McKusick 1983). Both diseases can be alternative causes of low GALNS activity depending on the sample type used. The possibility of MSD should be evaluated by measuring the activity of a second sulfatase such as arylsulfatase B (EC 3.1.6.12) or iduronate-2-sulfatase (EC 3.1.6.13). If a leukocyte sample or a DBS was used for GALNS analysis, ML II/III is not a concern as GALNS activity is not decreased by ML II/III in this specific sample type. However, if fibroblasts were used, ML II/III must be ruled out. This can be accomplished by measuring a second mannose-6-phosphate targeted enzyme such as β-galactosidase, arylsulfatase B, iduronate-2-sulfatase, β-hexosaminidase (EC 3.2.1.52), or α-iduronidase (EC 3.2.1.76) in fibroblasts.

Lastly, ruling out MPS IVB during enzyme analysis is also highly recommended. MPS IVA and B can present with very similar symptoms and can both cause elevated urinary KS (McKusick and Neufeld 1983). Therefore, β-galactosidase, the enzyme deficient in MPS IVB, is commonly tested in conjunction with GALNS. As mentioned previously, β-galactosidase activity in fibroblasts or plasma can also be used to rule out ML II/III (not necessary if leukocytes were used), allowing for the exclusion of two diseases with one measurement. Additional examples of enzymes that can be used to exclude more than one condition, including some additional alternative MPS disorders, when testing for MPS IVA are shown in Table 3.

Care should be taken when interpreting enzyme activity results. Reference ranges vary significantly depending on the units, sample type, and methodology used, as well as among laboratories (Table 4). Laboratories should clearly indicate reference ranges along with patient results and provide an interpretation in this context.

Molecular analysis, also known as mutation analysis, can be used to confirm enzyme activity results and facilitate genetic counseling of the family. The goal is to identify genetic mutations that result in decreased or absent GALNS enzyme activity. Because MPS IVA is a recessive disease, both GALNS alleles must contain a pathogenic mutation for a patient to be affected. In addition to pathogenic mutations, polymorphisms have also been reported in the GALNS gene (Tomatsu et al 2005a). When a novel mutation is identified, further investigation may be needed to determine whether or not it is pathogenic in nature.

The GALNS gene is located on chromosome 16q24.3 (Baker et al 1993; Masuno et al 1993), contains 14 exons (Nakashima et al 1994), and generates a 1566 nucleotide mRNA (Tomatsu et al 1991). A review of 148 unique mutations was published in 2005 (Tomatsu et al 2005a) and the identification of additional mutations continues at a steady pace to date. Mutation types reported in the Human Gene Mutation Database (HGMD) include missense, nonsense, splicing, small insertions and deletions, gross insertions and deletions, and complex rearrangements (Table 5). There is limited published data regarding mutation analysis of the GALNS gene, however the most common mutations published to date are reported to be missense mutations c.1156C>T (p.Arg386Cys), c.901G>T (p.Gly301Cys), and c.337A>T (p.Ile113Phe) (Tomatsu et al 2005a). However, their allelic frequencies are only 8.9, 6.8, and 5.7 %, respectively, demonstrating the allelic heterogeneity of MPS IVA (Tomatsu et al 2005a). Allelic frequencies are also very population-dependent. For example, the c.337A>T (p.Ile113Phe) mutation was actually found to be much more common (identified in 32 % of patients) in a cohort of United Kingdom/Eire patients (n = 89) evaluated at the Central Manchester University Hospitals in the United Kingdom (unpublished data).

Molecular analysis is typically performed using a blood sample. DNA in dried blood samples is very stable (Chaisomchit et al 2005) and, therefore, can be shipped easily. Specific collection cards are available for samples intended for DNA analysis; collection card selection should be as per receiving laboratory recommendations. It is important to note that blood samples from patients who have undergone a recent blood transfusion or a bone marrow/stem cell transplant at any time may give inaccurate results as these samples will likely be contaminated with donor DNA. These samples are not recommended for genotype analysis. If the initial biochemical diagnosis has been performed on fibroblasts, then the same cells could also be used as a source of patient DNA. Saliva samples and buccal swabs are additional acceptable sources of DNA.

In standard DNA sequencing-based methods, coding regions of the GALNS gene and small segments of immediately adjacent intronic regions are evaluated. Numerous MPS IVA mutations are believed to be “private” and found in only one individual or family (Tomatsu et al 2005a). Although many mutations have been reported in the literature, novel, unreported mutations are still commonly detected (Table 6). Unfortunately, conclusively determining whether a novel mutation is pathogenic may not always be possible. It is notable that some of the known pathogenic mutations in MPS IVA result in conservative amino acid changes, such as the common pathogenic mutation p.Ile113Phe, highlighting the difficulty in predicting the pathogenic nature of a novel mutation without associating enzyme activity testing results. Development of a global database of mutations with associated degree of compromised enzyme activity and phenotypic correlations could potentially facilitate improved dissemination of information and help reduce the number of mutations with unknown significance identified by GALNS sequencing.

While DNA sequencing-based methods identify missense mutations, nonsense mutations, and small insertions and deletions in the coding region of the gene, mutations that cause splicing alterations (Pajares et al 2012) or changes in copy number (Fukuda et al 1996) may be missed. If a pathogenic mutation is not found in each allele, evaluation of cDNA may be helpful in identifying intronic mutations causing splicing errors (Pajares et al 2012) and quantitative methods of molecular analysis can be used to identify mutations affecting copy number, such as large insertions and deletions (Fukuda et al 1996). A variety of methods are available including real-time PCR, comparative genomic hybridization (CGH), and multiplex ligation-dependent probe amplification (MLPA). Unfortunately, even when both sequencing and quantitative methods are used, some mutations, such as those in promoter regions, will still not be detectable. Undetectable mutations have been reported in the literature to occur in approximately 14 % of alleles (Tomatsu et al 2005a).

When two pathogenic mutations are successfully identified, it is important to confirm that they are on separate alleles (in trans) because more than one mutation may be present on the same allele (in cis). It is also important to identify any in cis mutations because their presence may impact future prenatal diagnoses. To confirm that the mutations are in trans, analyzing the parental DNA is recommended. If each parent has one of the mutations, they are likely present in trans in the patient. If both of the mutations are identified in one parent, the two identified mutations are likely in cis in the patient. In this case, the patient is either a carrier (not affected) or the mutation in the second allele has not yet been identified.

One notable exception to the mutations coming from both parents is a recently reported case of maternal uniparental isodisomy (Catarzi et al 2012). This case provides evidence for the need to perform parental testing in all cases, even when the patient is found to be homozygous for a known mutation. It is also possible that a change could occur de novo in the germ cell. These considerations highlight the importance of parental testing in being able to provide families with accurate genetic counseling regarding recurrence risk.

Ultimately, although molecular analysis results can confirm a diagnosis and allow for genetic counseling of the family, results can also be inconclusive and complicate the diagnostic process. Discussing the limitations of the analysis with the patient and their family prior to testing is recommended.