On the molecular nature of the Duarte variant of galactose-1-phosphate uridyl transferase (GALT)

Clinical and biochemical phenotypes, genotypes, and long-term outcomes of individuals with galactosemia type I from a single metabolic genetics center in Alberta

Background

Galactosemia type I is an autosomal recessive disorder of galactose metabolism due to galactose-1-phosphate uridyltransferase deficiency, encoded by GALT. To investigate the phenotypes, genotypes and long-term outcomes of galactosemia, we performed a retrospective cohort study in our center.

Methods

All individuals with galactosemia type I were included. We divided individuals into two groups to compare the outcomes of those treated symptomatically (SymX) and asymptomatically (AsymX). We reviewed electronic patient charts for clinical features, biochemical investigations, molecular genetic investigations, treatments, and outcomes.

Results

There were 25 individuals including classic (n = 17), clinical variant (n = 4), and biochemical variant (Duarte) galactosemia (n = 4). Twelve individuals were diagnosed symptomatically (SymX), and 9 individuals were diagnosed asymptomatically (AsymX). We did not include individuals with biochemical variant (Duarte) galactosemia into any of these groups. At the time of the diagnosis, conjugated hyperbilirubinemia was present in 83.3% of SymX group, whereas only 22% of AsymX group. SymX group had hepatomegaly (25%), failure to thrive (33.3%), cataract (16.7%) and sepsis (25%), whereas none of the individuals in the AsymX group had these clinical features. Fourteen variants in GALT were identified including pathogenic/likely pathogenic (n = 12), and likely benign/benign (n = 2) variants. The vast majority of individuals with classic and clinical variant galactosemia were treated with a galactose-lactose-free diet for life (n = 20/21). Intellectual disability was present in 54.5% of the SymX group, and in 37.5% of the AsymX group as a long-term outcome. Tremors were present 50% of the SymX group, and in 22% of the AsymX group as a long-term outcome. Although, intellectual disability and tremors seem to be less common in the AsymX group, there was no statistically significant difference between both groups. Primary ovarian insufficiency was present 50% of the SymX group, whereas in 20% of the AsymX group in post-pubertal females. We report a novel hypomorphic GALT variant (p.Ala303Ser) in one individual with clinical variant galactosemia. We also report an individual with clinical variant galactosemia with normal urine galactitol levels on a normal diet.

Conclusion

It seems that newborn screening and early administration of a galactose-lactose-free diet decreases the long-term galactosemia-associated complications but does not prevent them completely. It may be that not all individuals with clinical variant galactosemia may need a galactose-lactose-free diet. It is timely to find new therapeutic strategies that can reduce the frequency of late-onset complications in galactosemia.

Origins, distribution and expression of the Duarte-2 (D2) allele of galactose-1-phosphate uridylyltransferase

Duarte galactosemia is a mild to asymptomatic condition that results from partial impairment of galactose-1-phosphate uridylyltransferase (GALT). Patients with Duarte galactosemia demonstrate reduced GALT activity and carry one profoundly impaired GALT allele (G) along with a second, partially impaired GALT allele (Duarte-2, D2). Molecular studies reveal at least five sequence changes on D2 alleles: a p.N314D missense substitution, three intronic base changes and a 4 bp deletion in the 5′ proximal sequence. The four non-coding sequence changes are unique to D2. The p.N314D substitution, however, is not; it is found together with a silent polymorphism, p.L218(TTA), on functionally normal Duarte-1 alleles (D1, also called Los Angeles or LA alleles). The HapMap database reveals that p.N314D is a common human variant, and cross-species comparisons implicate D314 as the ancestral allele. The p.N314D substitution is also functionally neutral in mammalian cell and yeast expression studies. In contrast, the 4 bp 5′ deletion characteristic of D2 alleles appears to be functionally impaired in reporter gene transfection studies. Here we present allele-specific qRT–PCR evidence that D2 alleles express less mRNA in vivo than their wild-type counterparts; the difference is small but statistically significant. Furthermore, we characterize the prevalence of the 4 bp deletion in GG, NN and DG populations; the deletion appears exclusive to D2 alleles. Combined, these data strongly implicate the 4 bp 5′ deletion as a causal mutation in Duarte galactosemia and suggest that direct tests for this deletion, as proposed here, could enhance or supplant current tests, which define D2 alleles on the basis of the presence and absence of linked coding sequence polymorphisms.

Simultaneous amplification, detection, and analysis of common mutations in the galactose-1-phosphate uridyl transferase gene

Classic galactosemia is an autosomal recessive inherited error of galactose metabolism. It is caused by lack of galactose-1-phosphate uridyl transferase, an enzyme that is required to metabolize galactose-1-phosphate to uridine diphosphate galactose. The build up of galactose-1-phosphate is toxic at high levels and can damage the liver, brain, eyes, and other vital organs. Over 200 mutations have been identified in affected individuals. We describe an assay to identify nine target mutations or variants in the galactose-1-phosphate uridyl transferase gene, namely p.Q188R, p.S135L, p.K285N, p.L195P, p.T138M, p.Y209C, IVS2-2 A>G, p.L218L, and p.N314D. A single long-range PCR is followed by a multiplexed nucleotide extension assay (single nucleotide extension) and capillary electrophoresis to detect simultaneously all nine target mutations/variants. Fifty-four previously characterized samples (47 clinical samples and seven controls) gave a 100% concordance. We also report a nontarget novel mutation, p.L192X, and its profile using single nucleotide extension. This assay can complement the enzyme activity assay and identify familial mutations for testing additional family members.

Galactitol and galactonate in red blood cells of children with the Duarte/galactosemia genotype

We measured galactitol, galactonate, and galactose-1-phosphate in the red blood cell (RBC) to elucidate the biochemical phenotype of infants with a Duarte/galactosemia (D/G) genotype by isotope dilution GC/MS. The RBC galactonate, galactitol and Gal-1-P were quantified in 14 D/G newborns on a lactose containing formula or breast milk, eight D/G newborns on a galactose-free formula, and 18 D/G children between 1 and 2 years of age that were on a regular diet. The results were compared with those of non-galactosemic subjects of comparable age. In the D/G newborns on regular formula/breast milk, the levels of RBC galactitol, galactonate, and Gal-1-P were significantly higher than those of D/G newborns on diet treatment and non-galactosemic newborns. There was no difference in the levels of RBC galactitol, galactonate, and Gal-1-P between D/G newborns on a lactose-restricted diet and the control group. There appears to be two different responses to dietary galactose intake in D/G children. The first group of D/G children placed on a regular diet after a year of lactose restriction had higher RBC galactitol, galactonate levels than those of non-galactosemic children. The mean level of RBC galactonate was higher and the mean value of RBC galactitol was as high as that of galactosemic (G/G) patients on diet treatment. The second group of D/G children on a regular diet had normal levels of RBC galactitol and galactonate. The levels of RBC Gal-1-P were normal in both groups of D/G patients. The alternative pathway products may reflect galactose intake better than RBC Gal-1-P in D/G children.

Analysis of common mutations in the galactose-1-phosphate uridyl transferase gene: new assays to increase the sensitivity and specificity of newborn screening for galactosemia

Classical galactosemia is a genetic disease caused by mutations in the galactose-1-phosphate uridyl transferase (GALT) gene. Prospective newborn screening for galactosemia is routine and utilizes the universally collected newborn dried blood specimen on filter paper. Screening for galactosemia is achieved through analysis of total galactose (galactose and galactose-1-phosphate) and/or determining the activity of the GALT enzyme. While this approach is effective, environmental factors and the high frequency of the Duarte D2 mutation (N314D) does lead to false positive results. Using DNA derived from the original newborn dried blood specimen and Light Cycler technology a panel of five assays able to detect the four most frequently encountered classical galactosemia alleles (Q188R, S135L, K285N, and L195P) and the N314D Duarte variant mutation are described. The five assays are performed simultaneously using common conditions. Including DNA preparation, set-up, amplification, and analysis the genotype data for all five loci is obtained in less than 2 hours. The assays are easily interpreted and amenable to high-throughput newborn screening. Mutational analysis is useful to reduce false positive results, differentiate D/G mixed heterozygotes from classical galactosemia, and to clearly identify a very high percentage of those affected by classical galactosemia.

Real time PCR assays to detect common mutations in the biotinidase gene and application of mutational analysis to newborn screening for biotinidase deficiency

Biotinidase deficiency is an autosomal recessive disorder of biotin metabolism caused by defects in the biotinidase gene. Symptoms of biotinidase deficiency are resolved or prevented with oral biotin supplementation and as such newborn screening is performed to prospectively identify affected individuals prior to the onset of symptoms. Biotinidase deficiency is detected by determining the activity of the biotinidase enzyme utilizing the newborn dried blood spot and colorimetric end point analysis. While newborn screening by enzyme analysis is effective, external factors may compromise results of the enzyme analysis and difficulty is encountered in distinguishing between complete and partial enzyme deficiencies. In the United States, the four mutations most commonly associated with complete biotinidase deficiency are c98:d7i3, Q456H, R538C, and the double mutation D444H:A171T. Partial biotinidase deficiency is almost universally attributed to the D444H mutation. To more effectively distinguish between profound and partial biotinidase deficiency, a panel of assays utilizing real time PCR and melting curve analysis using Light Cycler technology was developed. Employing DNA extracted from the original dried blood specimens from newborns identified through prospective newborn screening as presumptive positive for biotinidase deficiency, the specimens were analyzed for the presence of the five common mutations. Using this approach it was possible to separate newborns with partial and complete deficiency from each other as well as from many of those with false positive results. In most cases it was also possible to correlate the genotype with the degree of residual enzyme activity present. In newborn screening for biotinidase deficiency, we have shown that the analysis of common mutations is useful in distinguishing between partial and complete enzyme deficiency as well as improving specificity. Combining biotinidase enzyme analysis with genotypic data also increases the sensitivity of screening for biotinidase deficiency and provides information useful to clinicians earlier than would otherwise be possible.

Characterization of a carbohydrate response element regulating the gene for human galactose-1-phosphate uridyltransferase

Human galactose-1-phosphate uridyltransferase (hGALT) is a central enzyme in the conserved pathway by which galactose is converted to energy, UDP-galactose and UDP-glucose. A natural mutation that deleted -119 to -116 bp (delGTCA) of the promoter decreased hGALT mRNA and enzyme activity and prompted analysis of hGALT gene regulation. Regulatory domains were identified by inspection and confirmed in a reporter system. Previous studies by others were confirmed that HepG2 cells grown in D-glucose increased hGALT enzyme activity and mRNA by 30%. We extended these observations by sequencing the promoter region and identifying a potential carbohydrate response element (ChoRE). The response to glucose rose to 190% when a plasmid construct containing a luciferase reporter and only the -165 bp region as a promoter was transfected into HepG2 and NIH:OVCAR-3. By contrast, fibroblasts transfected with the identical construct failed to respond to glucose. Within the -165 bp region there were two enhancer (E-box) motifs that encompassed the delGTCA mutation. The deletion diminished the positive regulatory response, but an additional GTCA repeat unexpectedly increased the response. Using this postulated ChoRE as a probe in electrophoretic mobility shift assays, multiple nuclear proteins bound and one was identified as a member of the basic/helix-loop-helix/leucine zipper enhancer-binding (b/HLH/LZ) family. Increased binding of proteins correlated with increased hGALT expression when the spacing between E-box motifs was enlarged but the carbohydrate response was dampened. When the 3(')E-box was mutated, b/HLH/LZ binding and gene expression were abolished. We conclude that the hGALT promoter region contains a ChoRE in which the spacing between and the sequence of its E-box motifs are critical. One nuclear protein of the b/HLH/LZ family is necessary, but not sufficient for the carbohydrate response.

A genetic factor for age-related cataract: identification and characterization of a novel galactokinase variant, "Osaka," in Asians

Galactokinase (GALK) deficiency is an autosomal recessive disorder characterized by hypergalactosemia and cataract formation. Through mass screening of newborn infants, we identified a novel and prevalent GALK variant (designated here as the "Osaka" variant) associated with an A198V mutation in three infants with mild GALK deficiency. GALK activity and the amount of immunoreactive protein in the mutant were both 20% of normal construct in expression analysis. The K(m) values for galactose and ATP-Mg(2+) in erythrocytes with homozygous A198V were similar to those of the healthy adult control subjects. A population study for A198V revealed prevalences of 4.1% in Japanese and 2.8% in Koreans, lower incidence in Taiwanese and Chinese, no incidence in blacks and whites from the United States, and a significantly high frequency (7.8%; P < .023) in Japanese individuals with bilateral cataract. This variant probably originated in Japanese and Korean ancestors and is one of the genetic factors that causes cataract in elderly individuals.

Functional analysis of the human galactose-1-phosphate uridyltransferase promoter in Duarte and LA variant galactosemia

Human galactose-1-phosphate uridyltransferase (hGALT) is an evolutionarily conserved enzyme central to D-galactose metabolism. The impairment of hGALT causes galactosemia. One missense mutation, an aspartate to asparagine substitution at amino acid 314 (N314D), impairs 50% activity in the homozygous state in some patients but gives near normal activity in others. The former condition is called Duarte (D) and the latter, Los Angeles (LA). The D allele is linked to hGALT polymorphisms including a deletion 5'to the translation start site (-119 to -116delGTCA), g1391G --> A and g1105G --> C. The LA allele is linked to a g1721C --> T transition. To investigate possible mechanisms for differences in hGALT activity between the D and LA alleles, we sequenced 3951 nucleotides of genomic DNA 5' to the hGALT translation start site. Using a dual-luciferase reporter system to express deletion constructs of the hGALT promoter, we noted both positive and negative regulatory regions. Two putative positive regulatory domains overlap with the naturally occurring -119 to -116delGTCA linked to Duarte. One is an E-box motif (CACGTG) at -117 to -112 bp. The second is an AP-1 motif (TCAGTCAG) at -124 to -119 bp. The delGTCA mutation confers reduced luciferase activity to transfected cell lines derived from human ovarian and liver neoplasms. Additionally, human lymphoblasts derived from patients with the Duarte allele have reduced GALT mRNA. We conclude that the human GALT gene is regulated in the first -165 bp of its promoter region by positive regulators of GALT gene expression. The -119 to -116delGTCA reduces hGALT transcription resulting in reduced GALT activity in the Duarte allele.

Molecular heterogeneity of classical and Duarte galactosemia: mutation analysis by denaturing gradient gel electrophoresis

Classical galactosemia is caused by one common missense mutation (Q188R) and by several rare mutations in the galactose-1-phosphate uridyltransferase (GALT) gene. The most common variant of GALT, the Duarte variant, occurs as two types, Duarte-1 (D-1) and Duarte-2 (D-2), both of which carry the sequence change N314D. D-1 increases, whereas D-2 decreases GALT activity. To study the molecular genetics of classical and Duarte galactosemia, we analyzed the GALT mutations in 30 families with classical galactosemia, in 10 families with the D-2 variant and in 3 individuals carrying the D-1 allele by denaturing gradient gel electrophoresis (DGGE). DGGE detected 59 of the 60 classical galactosemia alleles. Q188R accounted for 60%, K285N accounted for 28% of these alleles. Eight novel candidate galactosemia mutations were found. On all D-2 alleles N314D occurred in cis with two intronic sequence changes, on the D-1 alleles in cis with a neutral mutation in exon 7. We conclude that the mutations causing galactosemia are highly heterogeneous and that K285N is a second common galactosemia mutation in our population.

Classical galactosemia and mutations at the galactose-1-phosphate uridyl transferase (GALT) gene

Classical galactosemia is caused by a deficiency in activity of the enzyme galactose-1-phosphate uridyl transferase (GALT), which, in turn, is caused by mutations at the GALT gene. The disorder exhibits considerable allelic heterogeneity and, at the end of 1998, more than 150 different base changes were recorded in 24 different populations and ethnic groups in 15 countries worldwide. The mutations most frequently cited are Q188R, K285N, S135L, and N314D. Q188R is the most common mutation in European populations or in those predominantly of European descent. Overall, it accounts for 60-70% of mutant chromosomes, but there are significant differences in its relative frequency in individual populations. Individuals homoallelic for Q188R tend to have a severe phenotype and this is in keeping with the virtually complete loss of enzyme activity observed in in vitro expression systems. Globally, K285N is rarer, but in many European populations it can be found on 25-40% of mutant chromosomes. It is invariably associated with a severe phenotype. S135L is found almost exclusively in African Americans. In vitro expression results are discrepant, but some individuals carrying S135L appear to exhibit GALT activity in some tissues. Duarte 1 (or Los Angeles) and Duarte 2 (or Duarte) variants carry the same amino acid substitution, N314D, even though D1 is associated with increased erythrocyte GALT activity and D2 with reduced activity. N314D is in linkage disequilibrium with other base changes that differ on the D1 and D2 alleles. N314D does not impair GALT activity in in vitro expression systems. However, there are differences in the abundance of GALT protein in lymphoblastoid cells lines from D2 and D1 individuals. It is unclear whether the specific molecular changes that distinguish the D1 and D2 alleles account for the different activities. The considerable genetic heterogeneity documented to date undoubtedly contributes to the phenotypic heterogeneity that is observed in galactosemia. The additional effects of nonallelic variation and other constitutional factors on phenotypic variability remain to be elucidated.

The fundamental importance of human galactose metabolism: lessons from genetics and biochemistry

Cloning and characterization of all three human galactose-metabolic genes (GALK, GALT and GALE) has led to the identification of a number of mutations which are generally of the missense type in patients with galactosemia, an inborn error of metabolism. The predominance of missense mutations is interesting, considering the general importance of galactose metabolism for cellular energy production and proper modification of glycoproteins and glycolipids. Abnormalities in both of these macromolecules have been described in transferase-deficiency galactosemia, the most common and best-studied form of galactosemia. Thus, the parallel biochemical and molecular genetic analyses of human galactose metabolism are shedding light on this under-appreciated metabolic pathway that is critical for cellular energy production, modification of cellular macromolecules and normal human development.

Identification of an E689K substitution as the molecular basis of the human acid alpha-glucosidase type 4 allozyme (GAA*4)

We have identified the molecular basis of the GAA*4 allozyme as a G to A transition at nt2065 which predicts the substitution of glutamic acid by lysine at codon 689 (E689K). The conclusion that this change represents the molecular basis of the GAA*4 allozyme is based on 1) presence of the G2065A in homozygosity in a known GAA*4 homozygote, 2) transient expression studies showing normal enzyme activity expressed by cDNA containing the G2065A transition and 3) isoelectric focusing studies showing a more cathodal pattern for the expressed product as compared to the common GAA*1, analogous to the patterns seen in normal and known GAA*4 lymphoid cells.

Molecular characterization of Duarte-1 and Duarte-2 variants of galactose-1-phosphate uridyltransferase

The N314D polymorphism was found in two different alleles of the galactose-1-phosphate uridyltransferase (GALT) gene, Duarte-1 (D1) and Duarte-2 (D2). Although both variants have identical electrophoretic mobility and isoelectro-focusing points, the galactose-1-phosphate uridyltransferase (GALT) activity varies: D1 alleles showed 110-130% of the normal RBC activity, but D2 alleles only 40-50%. We found that D1 alleles also carried a silent mutation in exon 7 (L218L) in addition to N314D. In contrast, besides N314D, D2 alleles carried two regulatory mutations, G1105C and G1391A, in introns D and E, respectively. In normal and Q188R alleles none of the above four mutations coexisted. However, some galactosaemia alleles with mutations other than Q188R, such as W316X and E340X of exon 10, also carried the N314D mutation. The W316X alleles existed in cis with the intron mutations (G1105C and G1391A), whereas those with E340X are in cis with L218L. In all cases examined, the intron mutations were not found in D1 alleles and no D2 alleles had the silent mutation of L218L. These results suggest that the decrease in the GALT activity in D2 may be due to regulation of the GALT gene expression. The G1105C site may be critical to the function of erythroid transcription factor NF-E1, since it flanks the core consensus sequence for one of its binding sites. The G1391A mutation may affect another cis-acting regulatory sequence. Alternatively, both mutations may be involved in an aberrant splice processing, which possibly results in a low level of correctly spliced mRNA.

Molecular characterization of galactosemia (type 1) mutations in Japanese

We characterized two novel mutations of the galactose-1-phosphate uridyltransferase (GALT) gene in two Japanese patients with GALT deficiency and identified N314D and R333W mutations, previously found in Caucasians. One novel missense mutation was an G-to-A transition in exon 8, resulting in the substitution of arginine by histidine at the codon 231 (R231H). GALT activity of the R231H mutant construct was reduced to 15% of normal controls in a COS cell expression system. The other was a splicing mutation, an A-to-G transition at the 38th nucleotide in exon 3 (318A-->G), resulting in a 38-bp deletion in the GALT cDNA by activating a cryptic splice acceptor site. In seven Japanese families (14 alleles for classic form and one allele for Duarte variant) with GALT deficiency, the R231H and 318A-->G mutations were found only on both alleles of the proband. The N314D and R333W mutations were found on one allele each. The Q188R was prevalent in the United States but not in Japanese patients. The N314D mutation was associated with the Duarte variant in Japanese persons, as well as in the United States. We speculate that classic galactosemia mutations appear to differ between Japanese and Caucasian patients. Our limited data set on galactosemia mutations in Japanese suggests that the N314D GALT mutation encoding the Duarte variant arose before Asian and Caucasian people diverged and that classic galactosemia mutations arose and/or accumulated after the divergence of Asian and Caucasian populations.

Characteristics of women with a family history of ovarian cancer. I. Galactose consumption and metabolism

BACKGROUND

Galactose metabolism may be a risk factor for ovarian cancer based upon evidence that galactose causes ovarian failure and that ovarian cancer arises from premature ovarian failure. This study examines galactose-1-phosphate uridyl transferase (GALT) activity in women with a family history of ovarian cancer (FOC) to determine if low GALT activity occurs in women who are at risk for but in whom ovarian cancer has not yet developed.

METHODS

The authors studied 106 premenopausal women (FOC patients) with one primary or two second-degree relatives with ovarian cancer compared with 116 age matched control subjects without a family history of ovarian cancer (FOC controls). All women completed questionnaires and had blood drawn to measure GALT activity and genotype.

RESULTS

Mean erythrocyte GALT activity, in micromoles of hexose conversion per hour per gram of hemoglobin was 21.5 in FOC patients, significantly lower than the mean of 23.1 observed in FOC control subjects, (P = 0.001). FOC patients more frequently displayed the Duarte variant of galactosemia as detected by electrophoresis. In a subset of 87 patients and 113 control subjects for whom DNA was available, the allelelic frequency of the Duarte variant based upon molecular genetic detection of the N314D mutation that is associated with the Duarte variant was 15.5% among FOC cases compared with 7.5% among control subjects (P < 0.02). Galactose consumption did not differ between FOC patients and control subjects.

CONCLUSION

Galactose metabolism differs between women with and without a family history of ovarian cancer, suggesting that it may be a genetic risk factor for ovarian cancer, possibly mediated through oocyte toxicity from galactose.

Neonatal screening for cystic fibrosis using immunoreactive trypsinogen and direct gene analysis: four years' experience

OBJECTIVE

To assess the performance and impact of a two tier neonatal screening programme for cystic fibrosis based on an initial estimation of immunoreactive trypsinogen followed by direct gene analysis.

DESIGN

Four year prospective study of two tier screening strategy. First tier: immunoreactive trypsinogen measured in dried blood spot samples from neonates aged 3-5 days. Second tier: direct gene analysis of cystic fibrosis mutations (delta F508, delta I506, G551D, G542X, and R553X) in samples with immunoreactive trypsinogen concentrations in highest 1% and in all neonates with meconium ileus or family history of cystic fibrosis.

SETTING

South Australian Neonatal Screening Programme, Adelaide.

SUBJECTS

All 88,752 neonates born in South Australia between December 1989 and December 1993.

INTERVENTIONS

Neonates with two identifiable mutations were referred directly for clinical assessment and confirmatory sweat test; infants with only one identifiable mutation were recalled for sweat test at age 3-4 weeks. Parents of neonates identified as carriers of cystic fibrosis mutation were counselled and offered genetic testing.

MAIN OUTCOME MEASURES

Identification of all children with cystic fibrosis in the screened population.

RESULTS

Of 1004 (1.13%) neonates with immunoreactive trypsinogen > or = 99th centile, 912 (90.8%) had no identifiable mutation. 23 neonates were homozygotes or compound heterozygotes; 69 carried one identifiable mutation, of whom six had positive sweat tests. Median age at clinical assessment for the 29 neonates with cystic fibrosis was 3 weeks; six had meconium ileus and two had affected siblings. 63 neonates were identified as carriers of a cystic fibrosis mutation. Extra laboratory costs for measuring immunoreactive trypsinogen and direct gene analysis were $A1.50 per neonate screened.

CONCLUSION

This strategy results in early and accurate diagnosis of cystic fibrosis and performs better than screening strategies based on immunoreactive trypsinogen measurement alone.