Abstract
Purpose
Our previous study of Type II congenital smell loss patients revealed a statistically significant lower prevalence of an FY ( ACKR1 , formerly DARC ) haplotype compared to controls. The present study correlates this genetic feature with subgroups of patients defined by specific smell and taste functions.
Methods
Smell and taste function measurements were performed by use of olfactometry and gustometry to define degree of abnormality of smell and taste function. Smell loss was classified as anosmia or hyposmia (types I, II or III). Taste loss was similarly classified as ageusia or hypogeusia (types I, II or III). Based upon these results patient erythrocyte antigen expression frequencies were categorized by smell and taste loss with results compared between patients within the Type II group and published controls.
Results
Comparison of antigen expression frequencies revealed a statistically significant decrease in incidence of an Fy b haplotype only among patients with type I hyposmia and any form of taste loss (hypogeusia). In all other patient groups erythrocyte antigens were expressed at normal frequencies.
Conclusions
Data suggest that Type II congenital smell loss patients who exhibit both type I hyposmia and hypogeusia are genetically distinct from all other patients with Type II congenital smell loss. This distinction is based on decreased Fy b expression which correlated with abnormalities in two sensory modalities (hyposmia type I and hypogeusia). Only patients with these two specific sensory abnormalities expressed the Fy b antigen (encoded by the ACKR1 gene on the long arm of chromosome 1) at frequencies different from controls.
1
Introduction
Congenital smell loss is a condition in which patients are born with significantly deficient olfaction. We have defined two major forms of this abnormality : Type I congenital smell loss patients are distinguished from Type II congenital smell loss patients due to the former’s association with severe somatic, gonadal, and developmental abnormalities (e.g., Kallman syndrome ) and the latter’s lack of these clinical abnormalities . Although the character and clinical severity of the olfactory losses between Type I and II congenital smell loss are similar, Type I congenital smell loss is generally associated with anosmia (inability to detect or recognize any odorant) whereas Type II congenital smell loss patients exhibit various degrees of sensory loss including anosmia, type I hyposmia (ability to detect but inability to recognize any odorant), or type II hyposmia (decreased ability to detect and recognize olfactory stimuli) . Some Type II congenital smell loss patients also exhibit hypogeusia .
The etiology and genetic factors associated with Type I congenital smell loss have been described extensively . However, no genetic factors had been defined for Type II congenital smell loss despite earlier attempts . In an effort to explore a genetic abnormality present in patients with Type II congenital smell loss, we performed extensive erythrocyte surface antigen testing among these patients . By using olfactometry measurements only , we demonstrated that Type II congenital smell loss patients expressed the Fy b (“Duffy b”) antigen at a statistically significantly lower frequency than controls . The gene encoding the Duffy blood group system, formerly known as DARC , has recently been termed atypical chemokine receptor 1 ( ACKR1 ) .
We continued this previous work and included both smell and taste abnormalities to evaluate the association with the Fy a and Fy b variants of the ACKR1 gene. Five other blood group systems with a total of 16 variants were tested as controls. We hypothesized that the patients’ observed abnormalities may be genetic and present ACKR1 as a candidate gene region for further studies in clinically distinct patient cohorts or families who have Type II congenital smell loss with type I hyposmia and hypogeusia.
2
Materials and methods
2.1
Patients
Ninety-nine patients [aged 6–70; 30 ± 16 years (mean ± SD); 44 men and 55 women] with lifelong smell loss without associated somatic abnormalities or hypogonadism were previously described . Each patient exhibited normal growth, gonadal function, and development for their age without family history of olfactory or other sensory disorders in their offspring. Physical examination of each patient’s head and neck revealed no clinical abnormalities. Radiological studies of the brain revealed that patients exhibited the presence of one or more olfactory bulbs although in some patients they were of smaller size than in normal subjects .
2.2
Clinical assessment of taste and smell function
Comprehensive evaluation of the patients’ sensory function, including both smell and taste function, was performed in accordance with the olfactometry and gustometry protocols described previously .
Olfactometry and gustometry were defined by measurements of four characteristics of each sensory function: (i) measurements of detection thresholds (DT), which defined the presence of olfactory or gustatory receptors; (ii) recognition thresholds (RT), which defined the relationship between olfactory and gustatory receptors and the brain; (iii) magnitude estimation (ME), which defined the number of functioning olfactory and gustatory receptors; and (iv) hedonics (H), which defined the role of the brain, per se, in the recognition of olfactory and gustatory stimuli. Four olfactory stimuli were used to define these measurements: pyridine (a pungent odor), nitrobenzene (bitter-almond odor), thiophene (petroleum-like odor), and amyl acetate (banana-like odor). The technique to obtain these measurements was a forced choice, three stimuli, staircase previously defined in detail .
Based upon these measurements, four major types of olfactory loss were defined. These were defined as: anosmia (DT = 0, RT = 0, ME = 0, H = 0), type I hyposmia (DT > 0, RT = 0, ME = 0, H = 0), type II hyposmia (DT > 0 < normal, RT > 0 < normal, ME > 0 < normal, H > 0 < normal), and type III hyposmia (DT = normal, RT = normal, ME > 0 < normal, H > 0 < normal) . Three patients had anosmia, 79 patients had type I hyposmia, 17 patients had type II hyposmia, and none had type III hyposmia.
For taste, four categories of loss were also defined: ageusia, type I hypogeusia, type II hypogeusia, and type III hypogeusia (with characteristics similar to those described for olfaction). Tastants used were NaCl for salt, sucrose for sweet, HCl for sour, and urea for bitter as previously described . No patient had ageusia, 1 had type I hypogeusia, 50 had type II hypogeusia, 17 had type III hypogeusia, and 31 had normal taste function.
2.3
Immunohematology
Blood group antigens including Fy a and Fy b were tested with standard blood group serology methods as described previously .
2.4
Statistical analysis
Comparisons of patient erythrocyte antigen expression frequencies in each patient group were made with respect to the expected antigen frequencies of the general Caucasian population. These expected antigen frequencies were obtained from The Blood Group Antigen FactsBook (Third Edition) and Blood Groups and Red Cell Antigens .
Significance of difference was measured by non-cumulative (exact) binomial test with p < 0.05 considered significant. Calculations were performed in Microsoft Excel and confirmed with the Stat Trek online tool ( http://stattrek.com/online-calculator/binomial.aspx ).
The binomial test was deemed the most appropriate method to compare the number of instances of a binary phenomenon occurring within a tested group (e.g., the number of patients expressing a blood group antigen among all tested patients) to an expected probability of the phenomenon occurring (e.g., the frequency of individuals in the control population who express the antigen). Statistically significant deviations indicate that the tested group differs significantly from the control group with regard to the probability of the phenomenon occurring within each group.
Since multiple comparisons were made between the patient and the control groups, the Bonferroni–Holm method was applied to the threshold of significance to counteract familywise error . Because 8 statistical tests were done for Fy a and Fy b with a single-comparison significance threshold of p < 0.05, the Bonferroni–Holm method adjusted significance threshold became p < 0.00625 (0.05/8) for the test resulting in the lowest p value. If significance were determined for the first test, a significance threshold of p < 0.0071 (0.05/7) and p < 0.0083 (0.05/6) was applied for the next lowest p thresholds. Descriptive results for 5 other blood groups were shown for comparison. Because our study hypothesis pertained to Fy a and Fy b only, there was no need to correct for multiple statistical testing of the other 16 blood group antigens.
2
Materials and methods
2.1
Patients
Ninety-nine patients [aged 6–70; 30 ± 16 years (mean ± SD); 44 men and 55 women] with lifelong smell loss without associated somatic abnormalities or hypogonadism were previously described . Each patient exhibited normal growth, gonadal function, and development for their age without family history of olfactory or other sensory disorders in their offspring. Physical examination of each patient’s head and neck revealed no clinical abnormalities. Radiological studies of the brain revealed that patients exhibited the presence of one or more olfactory bulbs although in some patients they were of smaller size than in normal subjects .
2.2
Clinical assessment of taste and smell function
Comprehensive evaluation of the patients’ sensory function, including both smell and taste function, was performed in accordance with the olfactometry and gustometry protocols described previously .
Olfactometry and gustometry were defined by measurements of four characteristics of each sensory function: (i) measurements of detection thresholds (DT), which defined the presence of olfactory or gustatory receptors; (ii) recognition thresholds (RT), which defined the relationship between olfactory and gustatory receptors and the brain; (iii) magnitude estimation (ME), which defined the number of functioning olfactory and gustatory receptors; and (iv) hedonics (H), which defined the role of the brain, per se, in the recognition of olfactory and gustatory stimuli. Four olfactory stimuli were used to define these measurements: pyridine (a pungent odor), nitrobenzene (bitter-almond odor), thiophene (petroleum-like odor), and amyl acetate (banana-like odor). The technique to obtain these measurements was a forced choice, three stimuli, staircase previously defined in detail .
Based upon these measurements, four major types of olfactory loss were defined. These were defined as: anosmia (DT = 0, RT = 0, ME = 0, H = 0), type I hyposmia (DT > 0, RT = 0, ME = 0, H = 0), type II hyposmia (DT > 0 < normal, RT > 0 < normal, ME > 0 < normal, H > 0 < normal), and type III hyposmia (DT = normal, RT = normal, ME > 0 < normal, H > 0 < normal) . Three patients had anosmia, 79 patients had type I hyposmia, 17 patients had type II hyposmia, and none had type III hyposmia.
For taste, four categories of loss were also defined: ageusia, type I hypogeusia, type II hypogeusia, and type III hypogeusia (with characteristics similar to those described for olfaction). Tastants used were NaCl for salt, sucrose for sweet, HCl for sour, and urea for bitter as previously described . No patient had ageusia, 1 had type I hypogeusia, 50 had type II hypogeusia, 17 had type III hypogeusia, and 31 had normal taste function.
2.3
Immunohematology
Blood group antigens including Fy a and Fy b were tested with standard blood group serology methods as described previously .
2.4
Statistical analysis
Comparisons of patient erythrocyte antigen expression frequencies in each patient group were made with respect to the expected antigen frequencies of the general Caucasian population. These expected antigen frequencies were obtained from The Blood Group Antigen FactsBook (Third Edition) and Blood Groups and Red Cell Antigens .
Significance of difference was measured by non-cumulative (exact) binomial test with p < 0.05 considered significant. Calculations were performed in Microsoft Excel and confirmed with the Stat Trek online tool ( http://stattrek.com/online-calculator/binomial.aspx ).
The binomial test was deemed the most appropriate method to compare the number of instances of a binary phenomenon occurring within a tested group (e.g., the number of patients expressing a blood group antigen among all tested patients) to an expected probability of the phenomenon occurring (e.g., the frequency of individuals in the control population who express the antigen). Statistically significant deviations indicate that the tested group differs significantly from the control group with regard to the probability of the phenomenon occurring within each group.
Since multiple comparisons were made between the patient and the control groups, the Bonferroni–Holm method was applied to the threshold of significance to counteract familywise error . Because 8 statistical tests were done for Fy a and Fy b with a single-comparison significance threshold of p < 0.05, the Bonferroni–Holm method adjusted significance threshold became p < 0.00625 (0.05/8) for the test resulting in the lowest p value. If significance were determined for the first test, a significance threshold of p < 0.0071 (0.05/7) and p < 0.0083 (0.05/6) was applied for the next lowest p thresholds. Descriptive results for 5 other blood groups were shown for comparison. Because our study hypothesis pertained to Fy a and Fy b only, there was no need to correct for multiple statistical testing of the other 16 blood group antigens.
3
Results
Previously we reported that patients with Type II congenital smell loss have a statistically significant lower prevalence of an FY ( ACKR1 , formerly DARC ) haplotype compared to controls. Here, we explored the hypothesis that this FY haplotype may correlate with distinct subgroups of the same patient cohort.
Patients with hyposmia (which all patients exhibited) could be differentiated further by presence or absence of abnormalities in taste function. Patients with type I hyposmia could be subdivided into two groups, those with hypogeusia of any character (n = 53) and those with normal taste function (n = 26). Similarly, patients with type II hyposmia could also be further subdivided into two groups, those with hypogeusia (n = 12) and those with normal taste function (n = 5). All three patients with anosmia presented with hypogeusia.
3.1
Characterization of patient cohorts
The majority of patients with Type II congenital smell loss had type I hyposmia (~ 82%) as their olfactory abnormality. Patients with type II hyposmia comprised ~ 16% and anosmia accounted for the remaining ~ 2%.
Hypogeusia was present in ~ 67% of all patients and was present in this same proportion in patients with type I and type II hyposmia. Of Type II congenital smell loss patients with hypogeusia, ~ 1% presented with type I hypogeusia, ~ 74% presented with type II hypogeusia, and ~ 25% presented with type III hypogeusia.
3.2
Patients with type I hyposmia
For patients characterized with type I hyposmia alone, expression of the antigen Fy b occurred at a significantly decreased frequency when compared to the control population ( Table 1 ). Among these patients, those presenting with hypogeusia expressed the antigen Fy b at a significantly lower frequency compared to the control population. Of 53 patients tested for the Fy b antigen, 66% (n = 35) tested positive (versus 83% expected, p = 0.001). However, among patients with type I hyposmia, those exhibiting normal taste function expressed both tested antigens (Fy a and Fy b ) at frequencies that did not differ significantly from the control population. These results may indicate that a trait inflicting sensory losses of both smell and taste functions was associated with an allele of the ACKR1 gene known to express the Fy antigens.
