JP7653229B2 - How to assess the placebo effect - Google Patents

Description

The present invention relates to a method for evaluating the placebo effect on improving sleep in an individual, and a marker for this purpose.

In recent years, many people have been suffering from sleep problems. Insomnia leads to a state in which the quality of sleep is impaired, with symptoms such as difficulty falling asleep, shallow sleep, waking up early, and not feeling like one has slept enough. Poor sleep quality not only leads to serious illnesses such as myocardial infarction and cerebral infarction, but can also reduce work efficiency due to daytime sleepiness and potentially lead to serious accidents. Improving sleep quality is an important health issue. Many types of sleep-improving drugs have been developed with the aim of improving sleep quality.

Generally, the effectiveness of a drug is evaluated by comparing the effects of the test drug with a placebo that does not contain any active ingredients. However, when taking a placebo, symptoms may improve due to the psychological effects associated with taking the drug; this is called the placebo effect. If the placebo effect is large, it dilutes the effect of the test drug, which hinders accurate evaluation of the drug's efficacy. It is known that the placebo effect is particularly pronounced in cases of mental or nervous system disorders. Evaluation of sleep improvers is also usually performed by comparative testing with a placebo. However, because sleep is strongly affected by mental or nervous conditions, the placebo effect on sleep improvement tends to be relatively large, which hinders accurate evaluation of sleep improvers.

In recent years, in order to reduce the influence of the placebo effect on the efficacy of a test drug, placebo lead-in studies have been conducted in which subjects are administered only a placebo before a comparative study of a test drug and a placebo, and subjects who show a strong placebo effect are excluded from the study. However, even placebo lead-in studies such as those described above are unable to sufficiently reduce the placebo effect, and furthermore, placebo lead-in studies have drawbacks such as increased costs for the study, increased burden on subjects, and longer study periods (see Non-Patent Document 1). Non-Patent Document 2 reports that the score on the Life Orientation Test is associated with the placebo effect on improving sleep, and suggests that this be taken into consideration when evaluating the placebo effect.

Technologies have been developed to investigate the current and future physiological state of the human body by analyzing nucleic acids such as DNA and RNA in biological samples. Nucleic acids derived from living organisms can be extracted from tissues such as blood, body fluids, secretions, etc. Patent documents 1 and 2 describe isolating nucleic acids such as RNA derived from a subject's skin cells from surface skin lipids (SSL) and using them as samples for biological analysis.

International Publication No. 2018/008319 International Publication No. 2020/091044

Clin Invest, 2013, 3(9):823-833 J Psychosom Res, 2007, 62(5):563-570

The present invention provides a method for evaluating the placebo effect on improving sleep in an individual, and a marker for doing so.

In one aspect, the present invention provides a marker for evaluating a placebo effect on improving sleep in a subject, comprising at least one selected from the group consisting of nucleic acids encoding the following genes: TNK2, PPP1R3B, WDR33, HIF1AN, HSP90B1, AP2M1, GPR108, FCGR3B, DNTTIP1, C6orf1, FAM192A, TEN1, TMEM123, KAT5, DYNC1H1, CLTC, RAB11FIP4, SWAP70, CD52, TRIM28, LAMTOR4, PTGES3, FBP1, and SDF4, and proteins encoded by the genes.
In another aspect, the present invention provides a method of assessing the effect of a placebo on improving sleep in a subject, comprising measuring expression of said marker in said subject.
In another aspect, the present invention provides a method for selecting subjects who do not have a high placebo effect on improving sleep, the method comprising measuring expression of said marker in the subject.
In another aspect, the present invention provides a method for evaluating the effectiveness of a sleep improving agent, the method comprising evaluating the sleep improving effect of administration of the sleep improving agent in subjects who do not have a high placebo effect on sleep improvement selected by the selection method.

According to the present invention, the placebo effect on sleep improvement of individual subjects can be evaluated without actually having the subjects take a placebo. Furthermore, according to the present invention, subjects who show a high placebo effect can be excluded in advance from subjects in a test for evaluating a sleep improver, so that the influence of an excessive placebo effect in the evaluation of the sleep improver can be suppressed without conducting a test to reduce the influence of the placebo effect, such as a placebo lead-in test. Therefore, the present invention realizes an accurate evaluation of the effects of a sleep improver.

All patents, non-patent literature, and other publications cited herein are hereby incorporated by reference in their entirety.

The names of the genes disclosed herein follow the Official Symbols listed in NCBI ([www.ncbi.nlm.nih.gov/]).

In this specification, "skin surface lipids (SSL)" refers to the fat-soluble fraction present on the surface of the skin, and is sometimes called sebum. In general, SSL mainly contains secretions from exocrine glands such as sebaceous glands in the skin, and exists on the skin surface in the form of a thin layer that covers the skin surface.

In this specification, unless otherwise specified, "skin" is a general term for the area of the body surface including the epidermis, dermis, hair follicles, and tissues such as sweat glands, sebaceous glands, and other glands.

In this specification, "sleep improvement" preferably refers to an improvement in sleep quality, i.e., an improvement in overall satisfaction with sleep, not limited to sleep time. "Sleep improvement" can be evaluated, for example, by the OSA Sleep Questionnaire (Yamamoto Yukari et al.: Brain and Mental Medicine 10:401-409, 1999). The OSA Sleep Questionnaire evaluates the quality of sleep based on an individual's subjective symptoms, i.e., general lifestyle, physical and mental state before going to bed, and subjective quality of sleep upon waking. More specifically, the OSA Sleep Questionnaire evaluates the quality of sleep comprehensively based on five evaluation scales, the first to fifth factors. The first factor is an evaluation of sleepiness upon waking, the second factor is an evaluation of sleep onset and sleep maintenance, the third factor is an evaluation of dreaminess, the fourth factor is an evaluation of fatigue recovery, and the fifth factor is an evaluation of sleep time. Therefore, "sleep improvement" in this specification can refer to improvement in one or more of the five factors evaluated in the OSA sleep questionnaire, namely, sleepiness upon awakening, sleep onset and sleep maintenance, dreaming, fatigue recovery, and sleep time. Alternatively, "sleep improvement" in this specification is reflected in an increase in the average score of the five factors in the OSA sleep questionnaire. For example, the magnitude of the overall "sleep improvement" effect caused by the placebo effect can be expressed as the increase in the average score of the five factors in the OSA sleep questionnaire.

(1. Markers for assessing placebo effect on sleep improvement)
If the placebo effect on sleep improvement in each individual can be predicted, it is possible to exclude in advance those who show a high placebo effect from the subjects of a test for evaluating sleep-improving drugs. This also makes it possible to suppress the influence of an excessive placebo effect in the evaluation of sleep-improving drugs, thereby enabling an accurate evaluation of the effects of the sleep-improving drug.

The present inventors have found that the sensitivity of a subject to a placebo is reflected in the expression of a particular gene in the subject before administration of the placebo.
As shown in the examples below, a sample for gene expression analysis was collected in advance from a subject, and then the subject was given a placebo to examine its sleep improvement effect. As a result, the gene expression patterns in the samples collected in advance were different between the subject group that showed a high placebo effect and the subject group that showed a low placebo effect. The nucleic acids encoding the genes that showed different expression between the two groups and the proteins encoded by these genes can be used as markers for evaluating the subject's response to a placebo sleep improver, in other words, the sleep improvement effect of the placebo on the subject. For example, it is possible to evaluate the placebo effect on the subject's sleep improvement or select subjects who do not have a high placebo effect on sleep improvement using the expression of the nucleic acids encoding the genes or the proteins encoded by these genes as an index.

Thus, in one aspect, the present invention provides a marker for assessing the placebo effect on improving a subject's sleep. The marker for assessing the placebo effect on improving a subject's sleep provided by the present invention (hereinafter also referred to as the marker of the present invention) makes it possible to assess the placebo effect on improving sleep of the individual subject from whom the marker is derived, without actually administering a placebo to the subject.

The marker of the present invention may include at least one selected from the group consisting of nucleic acids encoding the 156 genes shown in Table 1 below, and proteins encoded by the genes (hereinafter also referred to as markers of Table 1). The marker of the present invention may be a nucleic acid encoding a gene shown in Table 1 below (nucleic acid marker), a protein encoded by the gene (protein marker), or a combination thereof, but is preferably a nucleic acid marker.

In one embodiment, the marker of the present invention comprises at least one selected from the group consisting of nucleic acids encoding genes shown in Tables 2 and 3 (hereinafter also referred to as nucleic acid markers of Tables 2 and 3, respectively) and proteins encoded by the genes (hereinafter also referred to as protein markers of Tables 2 and 3, respectively). The nucleic acid markers and protein markers in Table 2 may be collectively referred to as markers of Table 2, and the nucleic acid markers and protein markers in Table 3 may be collectively referred to as markers of Table 3. Preferably, the markers in Tables 2 and 3 are nucleic acid markers.
As shown in the Examples below, the 47 genes shown in Table 2 are highly expressed in the group with a high placebo effect, and the 33 genes shown in Table 3 are low expressed in the group with a high placebo effect. Therefore, the markers in Table 2 are positive markers whose expression levels are high in subjects with a high placebo effect on sleep improvement. On the other hand, the markers in Table 3 are negative markers whose expression levels are low in subjects with a high placebo effect on sleep improvement. In the present invention, either the former positive marker or the latter negative marker may be used, or both may be used in combination.

In another embodiment, the marker of the present invention comprises at least one selected from the group consisting of nucleic acids encoding the 100 genes shown in Table 4 (hereinafter also referred to as nucleic acid markers of Table 4) and proteins encoded by the genes (hereinafter also referred to as protein markers of Table 4). The nucleic acid markers and protein markers of Table 4 may be collectively referred to as the markers of Table 4. Preferably, the markers of Table 4 are nucleic acid markers.
As shown in the Examples described below, the genes shown in Table 4 are those that correspond to the top 100 in variable importance based on the Gini coefficient when an attempt was made to extract feature genes and construct a prediction model using Random Forest as a machine learning algorithm, with data on the expression levels in subjects as explanatory variables and whether the placebo effect on improving sleep is high or low as the objective variable.

In a preferred embodiment, the marker of the present invention comprises at least one selected from the group consisting of nucleic acids encoding the 24 genes shown in Table 5 (hereinafter also referred to as nucleic acid markers of Table 5) and proteins encoded by the genes (hereinafter also referred to as protein markers of Table 5). The nucleic acid markers and protein markers of Table 5 may be collectively referred to as the markers of Table 5. Preferably, the markers of Table 5 are nucleic acid markers.
The genes shown in Table 5 are genes commonly included in Tables 2 or 3 and 4.

In a preferred embodiment, the markers of the present invention may further comprise at least one selected from the group consisting of the markers in Table 1 (but other than the markers in Table 5) in addition to at least one selected from the group consisting of the markers in Table 5. For example, the markers of the present invention may further comprise at least one selected from the group consisting of the markers in Table 2, Table 3, or Table 4 (but other than the markers in Table 5) in addition to at least one selected from the group consisting of the markers in Table 5.
In this case, the at least one selected from the group consisting of the markers in Table 5 is preferably all 24 of the markers shown in Table 5.

In another preferred embodiment, all of the nucleic acid markers in Tables 2 and 3, or all of the protein markers in Tables 2 and 3, are used in combination as markers of the present invention.
In another preferred embodiment, all of the nucleic acid or protein markers in Table 4 are used in combination as markers of the present invention.
In another preferred embodiment, all of the nucleic acid or protein markers in Table 5 are used in combination as markers of the present invention.
In another preferred embodiment, all of the nucleic acid or protein markers in Table 1 are used in combination as markers of the present invention.

The markers of the present invention can be prepared from biological samples collected from subjects, such as cells, body fluids (blood, etc.), urine, secretions (saliva, SSL, etc.), etc., according to standard methods. For example, commercially available kits can be used to prepare nucleic acids or proteins from biological samples. Preferably, the markers of the present invention are nucleic acid markers, and preferred examples of nucleic acids prepared from biological samples include genomic DNA and mRNA.

Examples of subjects from which biological samples containing the markers of the present invention are collected include those who can be subjects in evaluation tests of sleep-improving drugs, such as healthy individuals who do not experience problems with sleep, as well as individuals with insomnia symptoms and individuals wishing to improve the quality of their sleep.

More preferably, the marker of the present invention is a nucleic acid or protein, even more preferably mRNA, prepared from the SSL of a subject. The site of the skin from which SSL is collected may be any site of the body, such as the head, face, neck, trunk, hands and feet, and preferably a site with high sebum secretion, such as the skin of the face.

To collect SSL from the skin of a subject, any means used for recovering or removing SSL from the skin can be used. Preferably, an SSL absorbent material, an SSL adhesive material, or an instrument for scraping SSL from the skin, as described below, can be used. The SSL absorbent material or SSL adhesive material is not particularly limited as long as it has an affinity for SSL, and examples thereof include polypropylene and pulp. More detailed examples of procedures for collecting SSL from the skin include a method of absorbing SSL into a sheet-like material such as oil blotting paper or oil blotting film, a method of adhering SSL to a glass plate or tape, and a method of scraping SSL off and collecting it with a spatula, scraper, etc. In order to improve the adsorption of SSL, an SSL absorbent material that has been previously impregnated with a highly lipid-soluble solvent may be used. On the other hand, it is preferable that the SSL absorbent material contains a low content of highly water-soluble solvents and water, since the adsorption of SSL is inhibited if the SSL absorbent material contains highly water-soluble solvents or water. It is preferable to use SSL absorbent material in a dry state.

The collected SSL may be used immediately in the nucleic acid or protein extraction step described below, or may be stored until it is used in the nucleic acid or protein extraction step. When stored, the SSL is preferably stored under low temperature conditions. The temperature conditions for storing the SSL may be 0° C. or lower, and are preferably −20±20° C. to −80±20° C., more preferably −20±10° C. to −80±10° C., even more preferably −20±20° C. to −40±20° C., even more preferably −20±10° C. to −40±10° C., even more preferably −20±10° C., and even more preferably −20±5° C. The period for storing the SSL is not particularly limited, but is preferably 12 months or less, for example, 6 hours or more and 12 months or less, more preferably 6 months or less, for example, 1 day or more and 6 months or less, even more preferably 3 months or less, for example, 3 days or more and 3 months or less.

For the extraction of nucleic acids or proteins from the collected SSL, methods that are commonly used for the extraction or purification of nucleic acids or proteins from biological samples can be used. Examples of nucleic acid extraction or purification methods include the phenol/chloroform method, the AGPC (acid guanidinium thiocyanate-phenol-chloroform extraction) method, or methods using columns such as TRIzol (registered trademark), RNeasy (registered trademark), and QIAzol (registered trademark), methods using special silica-coated magnetic particles, methods using Solid Phase Reversible Immobilization magnetic particles, and extraction using commercially available RNA extraction reagents such as ISOGEN. For protein extraction or purification, commercially available protein extraction reagents such as QIAzol Lysis Reagent (Qiagen) can be used.

2. Methods for assessing placebo effects on sleep improvement
In another aspect, the present invention provides a method for evaluating the placebo effect on improving sleep in a subject using the marker of the present invention described in 1 above. In this method, the placebo effect on improving sleep in an individual subject is evaluated based on the expression of the marker of the present invention in the individual subject. In yet another aspect, the present invention provides a method for selecting a subject for whom the placebo effect on improving sleep is not high using the marker of the present invention described in 1 above. In this method, a subject for whom the placebo effect on improving sleep is not high is selected based on the expression of the marker of the present invention in the individual subject. Hereinafter, the above methods may be collectively referred to as the method of the present invention.

2.1 Analysis of Marker Expression
The subject to be subjected to the method of the present invention is the same as the subject from which the biological sample containing the above-mentioned marker of the present invention is collected. The subject is in a state where placebo is not administered. Therefore, the method of the present invention makes it possible to evaluate the placebo effect on the improvement of sleep of the subject without actually administering a placebo to the subject, or to select a subject who does not have a high placebo effect on improving sleep.

In a preferred embodiment, the method of the present invention comprises measuring the expression of a marker of the present invention in a biological sample collected from a subject. The type of the biological sample is as described above, preferably SSL. In one embodiment, the method of the present invention may further comprise collecting the SSL of the subject. The procedure for collecting the SSL and the procedure for extracting the marker from the SSL are as described above.

The expression of the markers of the present invention can be measured according to a method for quantifying nucleic acids or proteins commonly used in the field. For example, the expression level of a nucleic acid marker may be measured according to a procedure for gene expression analysis commonly used in the field. Examples of gene expression analysis techniques include methods for quantifying DNA or its amplification products by real-time PCR, multiplex PCR, microarray, sequencing, chromatography, etc. When the nucleic acid is RNA, the RNA may be converted to cDNA by reverse transcription and then quantified by the above method. The expression level of a protein marker can be measured using a protein quantification method commonly used in the field, such as ELISA, immunostaining, fluorescence, electrophoresis, chromatography, mass spectrometry, etc. The calculated expression level of a marker may be an expression level based on the absolute amount of the target marker in a biological sample, or may be a relative expression level to the expression level of other standard substances or the total nucleic acid or protein.

Preferably, the marker of the present invention used in the method of the present invention is SSL-derived mRNA. The RNA is converted to cDNA by reverse transcription, and then the cDNA is subjected to PCR, and the resulting reaction product is purified. The reverse transcription may be performed using a primer targeted to the specific RNA to be analyzed, but for more comprehensive nucleic acid storage and analysis, it is preferable to use a random primer. A general reverse transcriptase or a reverse transcription reagent kit can be used for the reverse transcription. Preferably, a highly accurate and efficient reverse transcriptase or reverse transcription reagent kit is used, and examples thereof include M-MLV Reverse Transcriptase and its variants, or commercially available reverse transcriptases or reverse transcription reagent kits, such as the PrimeScript (registered trademark) Reverse Transcriptase series (Takara Bio Inc.) and the SuperScript (registered trademark) Reverse Transcriptase series (Thermo Scientific). SuperScript (registered trademark) III Reverse Transcriptase, SuperScript (registered trademark) VILO cDNA Synthesis kit (both from Thermo Scientific) and the like are preferably used. In PCR of the obtained cDNA, a primer pair targeting a specific DNA to be analyzed may be used to amplify only the specific DNA, or multiple primer pairs may be used to amplify multiple DNAs. Preferably, the PCR is multiplex PCR. Multiplex PCR is a method in which multiple gene regions are amplified simultaneously by simultaneously using multiple primer pairs in a PCR reaction system. Multiplex PCR can be performed using a commercially available kit (e.g., Ion AmpliSeq Transcriptome Human Gene Expression Kit; Life Technologies Japan, Inc., etc.).

By adjusting the reaction conditions of the reverse transcription and PCR, the yield of the PCR reaction product is further improved, and the accuracy of the analysis using the same is further improved. Preferably, in the extension reaction in the reverse transcription, the temperature is preferably adjusted to 42°C ± 1°C, more preferably 42°C ± 0.5°C, and even more preferably 42°C ± 0.25°C, while the reaction time is preferably adjusted to 60 minutes or more, more preferably 80 to 120 minutes. Also preferably, the temperature of the annealing and extension reaction in the PCR is preferably 62°C ± 1°C, more preferably 62°C ± 0.5°C, and even more preferably 62°C ± 0.25°C. Therefore, in the PCR, the annealing and extension reactions are preferably performed in one step. The time of the annealing and extension reaction steps can be adjusted depending on the size of the DNA to be amplified, etc., but is preferably 14 to 18 minutes. The conditions of the denaturation reaction in the PCR can be adjusted depending on the DNA to be amplified, but is preferably 95 to 99°C for 10 to 60 seconds. Reverse transcription and PCR at the temperatures and times described above can be carried out using a thermal cycler commonly used for PCR.

The purification of the reaction products obtained by the PCR is preferably carried out by size separation of the reaction products. By size separation, the target PCR reaction products can be separated from primers and other impurities contained in the PCR reaction solution. Size separation of DNA can be carried out, for example, by a size separation column, a size separation chip, or magnetic beads that can be used for size separation. A preferred example of magnetic beads that can be used for size separation is Solid Phase Reversible Immobilization (SPRI) magnetic beads such as Ampure XP.

The purified PCR reaction product may be subjected to further processing necessary for subsequent quantitative analysis. For example, for DNA sequencing, the purified PCR reaction product may be prepared in an appropriate buffer solution, the PCR primer region contained in the PCR-amplified DNA may be cleaved, or an adapter sequence may be further added to the amplified DNA. For example, the purified PCR reaction product may be prepared in a buffer solution, the PCR primer sequence may be removed from the amplified DNA, and adapter ligation may be performed, and the resulting reaction product may be amplified as necessary to prepare a library for quantitative analysis. These operations can be performed, for example, using the 5×VILO RT Reaction Mix included in the SuperScript® VILO cDNA Synthesis kit (Life Technologies Japan, Inc.), and the 5×Ion AmpliSeq HiFi Mix and Ion AmpliSeq Transcriptome Human Gene Expression Kit (Life Technologies Japan, Inc.) according to the protocol included with each kit. For sequencing, a next-generation sequencer (e.g., Ion S5/XL system, Life Technologies Japan, Inc.) is preferably used. RNA expression can be quantified based on the number of reads generated by sequencing (read count).

2.2 Evaluation of placebo effect based on marker expression levels
In one embodiment of the method of the present invention, the expression level of a marker of the present invention derived from each subject (a marker of the present invention contained in a biological sample collected from the subject) is used as an index for evaluating the placebo effect on improving sleep in the subject (e.g., whether the placebo effect is high or low), or for selecting subjects for whom the placebo effect is not high.

As described above, the markers in Tables 2 and 3 are markers whose expression levels vary depending on the placebo effect on sleep improvement in the subject. More specifically, the markers in Table 2 are positive markers whose expression levels are high in groups with a high placebo effect on sleep improvement, while the markers in Table 3 are negative markers whose expression levels are low in groups with a high placebo effect on sleep improvement. Therefore, the expression levels of the positive or negative markers can be used as indicators to evaluate the level of the placebo effect on sleep improvement in the subject (e.g., whether the placebo effect is high or low), or to determine whether the subject is a person for whom the placebo effect on sleep improvement is high.

In this embodiment, one or more markers selected from the positive markers and the negative markers in a biological sample collected from a subject can be used as target markers to be used as indicators for the above-mentioned evaluation or selection. In the method of the present invention, either one of the positive markers and the negative markers may be used as a target marker, or two or more selected from the positive markers and the negative markers may be used in combination as a target marker. Alternatively, a nucleic acid marker or a protein marker consisting of all of the positive markers and the negative markers may be used in combination as a target marker.

For example, by measuring the expression level of a target marker and comparing the measured expression level of the target marker with a reference value, it is possible to determine whether or not the subject is a person for whom the placebo effect on improving sleep is high (in other words, whether or not the placebo effect on improving sleep in the subject is high).
The reference value may be a predetermined value, for example, a statistical value (e.g., average value) of the expression level of the target marker in individuals who have a high placebo effect on improving sleep (high placebo effect group) or individuals who have a low placebo effect on improving sleep (low placebo effect group). When multiple types of nucleic acids or proteins are used as target markers, it is preferable to determine the reference value for each nucleic acid or protein.

In one example, if the expression level of the positive marker derived from a subject is equal to or lower than that of the low placebo effect group, the subject is judged to be a person who does not have a high placebo effect on improving sleep (in other words, the placebo effect on improving the subject's sleep is not high). Alternatively, if the expression level of the positive marker derived from a subject is lower than that of the high placebo effect group, the subject is judged to be a person who does not have a high placebo effect on improving sleep (in other words, the placebo effect on improving the subject's sleep is not high).
On the other hand, if the expression level of the positive marker is higher than that of the low placebo effect group or equal to or higher than that of the high placebo effect group, the subject is not judged to be one in which the placebo effect on improving sleep is not high (in other words, the placebo effect on improving sleep for the subject is not high).

In another example, if the expression level of the negative marker derived from the subject is equal to or higher than that of the low placebo effect group, the subject is judged to be a person for whom the placebo effect on sleep improvement is not high (in other words, the placebo effect on the subject's sleep improvement is not high). Alternatively, if the expression level of the negative marker derived from the subject is higher than that of the high placebo effect group, the subject is judged to be a person for whom the placebo effect on sleep improvement is not high (in other words, the placebo effect on the subject's sleep improvement is not high).
On the other hand, if the expression level of the negative marker is lower than that of the low placebo effect group or equal to or lower than that of the high placebo effect group, the subject is not judged to be one in whom the placebo effect on improving sleep is not high (in other words, the placebo effect on improving sleep for that subject is not high).

The difference between the expression level of a marker derived from a subject and a reference value can be evaluated, for example, by whether or not the two are statistically significantly different. When multiple markers are used as target markers, the expression level of each target marker is compared with the reference value, and a certain percentage, for example, 50% or more, preferably 70% or more, more preferably 90% or more, and even more preferably 100% of the marker expression levels are examined to see if they differ from the reference value, thereby evaluating the placebo effect on improving the subject's sleep.

(2.3 Evaluation of placebo effect based on predictive models)
In another embodiment of the method of the present invention, the placebo effect (e.g., whether the placebo effect is high or low) on improving sleep in the subject is evaluated based on a prediction model constructed using data on the expression of the markers of the present invention (the markers of the present invention contained in a biological sample collected from the subject) (hereinafter referred to as an expression profile), or a subject in which the placebo effect is not high is selected. Examples of the expression profile include data on expression levels such as the number of sequencing reads.

For example, by using machine learning with the expression profiles of one or more markers (nucleic acids or proteins) obtained from each individual in a teacher sample population (e.g., a population including a high placebo effect group and a low placebo effect group) as explanatory variables, and data on the level of the placebo effect on sleep improvement in each individual in the population (e.g., whether the placebo effect is high or low) as the objective variable, a predictive model (e.g., a discriminant) can be constructed to evaluate the placebo effect on sleep improvement in any subject.
The constructed predictive model can be used to evaluate the placebo effect on sleep improvement in individual subjects (e.g., whether the placebo effect is high or low) or to select subjects who do not have a high placebo effect.

In this specification, "feature" is synonymous with "explanatory variable" in machine learning. In this specification, markers whose expression profiles are used as explanatory variables (features) in machine learning may be referred to as "feature marker(s)." Furthermore, when a feature marker is a nucleic acid marker, it may be referred to as a feature gene.

The feature marker (group) used in this embodiment may be at least one selected from the group consisting of nucleic acids encoding the genes shown in Table 1 and proteins encoded by the genes. The expression profile of the feature marker may be an absolute value or a relative value, or may have been normalized. When multiple markers are used as the feature marker group, for example, multiple markers that are highly correlated with the placebo effect on improving sleep can be selected from the markers of the present invention, and each of their expression profiles can be used as an explanatory variable.

In a preferred embodiment, all of the nucleic acid markers in Tables 2 and 3, or all of the protein markers in Tables 2 and 3, are combined and used as a group of feature markers. In another preferred embodiment, all of the nucleic acid or protein markers in Table 4 are combined and used as a group of feature markers. In another preferred embodiment, all of the nucleic acid or protein markers in Table 5 are combined and used as a group of feature markers. In another preferred embodiment, all of the nucleic acid or protein markers in Table 1 are combined and used as a group of feature markers.

In a preferred embodiment, expression profiles of feature markers (groups) in subjects with a high placebo effect on sleep improvement (high placebo effect group) and subjects with a low placebo effect on sleep improvement (low placebo effect group) are used as teacher samples for machine learning. Using the teacher samples, a discriminant (prediction model) for dividing subjects into a high placebo effect group and a low placebo effect group is constructed. The expression profile of the feature markers (groups) can be used as an explanatory variable used in constructing the discriminant. For example, a variable indicating whether the subject from whom the feature markers (groups) are derived is a high placebo effect group or a low placebo effect group can be used as an objective variable. Based on the constructed discriminant, a cutoff value for discriminating between the high placebo effect group and the low placebo effect group can be obtained. Next, the expression profile of the feature markers (groups) derived from the subject for whom the placebo effect is to be examined is measured, the measured value obtained is substituted into the discriminant, and the result obtained from the discriminant is compared with the cutoff value to discriminate whether the subject is a high placebo effect group or a low placebo effect group.

For example, a statistically significant difference between two groups to be discriminated (high placebo effect group and low placebo effect group), for example, a marker whose expression profile varies significantly between the two groups, can be used as a feature marker (group), and the expression profile can be used as an explanatory variable. Alternatively, the feature marker (group) can be extracted using a known algorithm such as a machine learning algorithm. For example, the feature marker (group) can be extracted using the variable importance in the random forest shown below, the "Boruta" package in the R language, or the like.

Alternatively, when using the expression profile of a marker to construct a predictive model, the data may be compressed by dimensionality reduction as necessary before constructing the predictive model. For example, multiple markers are extracted from the nucleic acid markers that code for the group of genes shown in Table 1. Next, principal component analysis is performed on the expression profile of the extracted markers. A predictive model for determining whether a subject is in the high placebo effect group or the low placebo effect group can be constructed by machine learning using one or more principal components calculated by the principal component analysis as explanatory variables and a variable indicating whether the subject from which the explanatory variable is derived is in the high placebo effect group or the low placebo effect group as the objective variable.

The algorithm used in constructing the predictive model can be a known algorithm such as an algorithm used in machine learning. Examples of machine learning algorithms include, but are not limited to, a linear regression model, Lasso regression, Random Forest, Neural Net, a support vector machine with a linear kernel (SVM (linear)), and a support vector machine with an rbf kernel (SVM (rbf)). Random Forest is preferred as the algorithm used in the present invention.

The prediction model constructed is input with validation data to calculate a prediction value, and the model in which the prediction value best matches the actual measurement value, for example, the model with the highest accuracy of the prediction value relative to the actual measurement value, or the model with the smallest root mean square error (RMSE) of the difference between the prediction value and the actual measurement value, can be selected as the optimal model. By inputting the expression profile of the feature marker(s) actually measured from the subject into the constructed prediction model, the placebo effect on sleep improvement in the subject (for example, whether the placebo effect is high or low) can be evaluated. Subjects who are judged to have a low placebo effect (or subjects who are not judged to have a high placebo effect) can be selected as subjects for whom the placebo effect on sleep improvement is not high.

(3. Placebo effect evaluation kit for improving sleep)
In a further aspect, the present invention provides a kit for evaluating the placebo effect on improving sleep in a subject according to the method of the present invention described in 2. above. In one embodiment, the kit of the present invention includes a reagent or instrument for measuring the expression of the marker of the present invention described above. For example, the kit of the present invention may include a reagent for amplifying or quantifying the nucleic acid marker of the present invention (e.g., reverse transcriptase, a PCR reagent, an adapter sequence for sequencing, etc.). Preferably, the kit of the present invention includes an index or guidance for evaluating the expression of the marker of the present invention. For example, the kit of the present invention may include guidance explaining the relationship between the increase or decrease in the expression level of each marker and the degree of the placebo effect, or guidance explaining the reference value of the expression level of each marker for evaluating the placebo effect, or guidance on a discriminant based on a prediction model and a feature marker to be input thereto. The kit of the present invention may further include a biological sample collection device (e.g., the above-mentioned SSL absorbent material or SSL adhesive material), a reagent for extracting the marker of the present invention from a biological sample (e.g., a nucleic acid purification reagent), and the like.

4. Methods for assessing the efficacy of sleep-improving agents
In a further aspect, the present invention provides a method for evaluating the efficacy of a sleep improver. The method includes evaluating the sleep-improving effect of administering the sleep improver in a subject. The subject is selected by the method of the present invention described in 2. above, and is not highly susceptible to the placebo effect on sleep improvement.

The subject is administered the sleep improver according to the general methodology of sleep improver evaluation studies. The sleep improver may be administered to the subject in a blinded manner. When conducting a comparative study with a placebo, the subjects may be divided into two groups, one of which is administered the sleep improver and the other a placebo. Alternatively, the subjects may be administered the sleep improver and a placebo alternately according to the general methodology of a crossover study. The subject's sleep state is also examined before and after administration of the sleep improver (or placebo) according to the general methodology of a sleep improver evaluation study.

The sleep-improving agent administered to the subject may be in the form of a medicine, a quasi-drug, or a food. Such foods include foods and supplements that are based on the concept of improving the quality of sleep and that are labeled as such as necessary. Examples of the form of the medicine (including quasi-drugs) include oral administration forms such as tablets, capsules, granules, powders, syrups, etc., and parenteral administration forms such as injections, suppositories, inhalants, etc., and are preferably oral administration forms. Examples of the form of the food include soft drinks, tea drinks, coffee drinks, fruit juice drinks, carbonated drinks, jellies, wafers, biscuits, bread, noodles, sausages, and other foods and nutritional foods, as well as nutritional supplement compositions in the same form as the above-mentioned oral administration preparations (tablets, capsules, syrups, etc.). The sleep-improving agent administered to the subject is preferably a food, and more preferably a beverage or supplement.

The placebo administered to the subject does not contain the active ingredient in the sleep-improving agent being evaluated, but can be appropriately manufactured to have the same appearance, flavor, and dosage form.

The sleep-improving effect of a sleep-improving agent on a subject administered the agent can be evaluated according to general methods for evaluating sleep-improving agents. For example, the effectiveness of the sleep-improving agent can be evaluated based on the results of an investigation into the subject's sleep state before and after administration of the sleep-improving agent to be evaluated. Alternatively, the effectiveness of the sleep-improving agent can be evaluated based on expression information of sleep-related factors in the subject's body obtained by genetic analysis, biochemical testing, etc.

If a comparison or crossover study with a placebo is conducted, the sleep-improving effect of administering the placebo is evaluated in subjects who received the placebo using the same procedure as described above. If the sleep-improving effect of administering the sleep-improving agent is greater than the sleep-improving effect of administering the placebo, the sleep-improving agent can be evaluated as effective.

The present invention will be described in more detail below with reference to examples, but the present invention is not limited to these.

Example 1: Construction of a placebo effect prediction model for sleep improvement using expression-varying genes 1) Study participants Study participants were recruited under the pretext of an evaluation study of a sleep-improving drug. Among the applicants, 103 healthy men and women in their 30s to 60s who were dissatisfied with their sleep and who answered "slightly bad" or "very bad" out of four levels of "very good,""slightlygood,""slightlybad," and "very bad" to the question, "How would you rate your overall quality of sleep over the past month?" were selected as study participants.

2) Collection of Sebum RNA Before the placebo ingestion test, sebum containing RNA was collected from the entire face of the test participants using an oil blotting film (5 x 8 cm, made of polypropylene, 3M). The oil blotting film was transferred to a glass vial and stored at -80°C until it was used for RNA extraction.

3) Placebo Intake Test As a placebo for the sleep improver, sugar-coated tablets (350 mg/tablet) mainly composed of maltitol and cellulose, which do not contain any materials that show a sleep-improving effect, were used. The placebo was given to the test participants without any information on whether it contained a sleep-improving active ingredient, and the participants were asked to take four placebo tablets once a day for one week, approximately 30 to 60 minutes before going to bed. The test participants were asked to answer the OSA Sleep Questionnaire (Yamamoto Yukari et al.: Brain and Mental Medicine 10:401-409, 1999), which is a psychological scale that evaluates sleep reflection at the time of waking, on the morning of the day the intake started and the morning after one week of intake. Based on the change in the OSA Sleep Questionnaire score (average score of factors 1 to 5) before and after intake, the effect of taking the placebo on the quality of sleep of the test participants was evaluated. As a result, it was confirmed that the degree of improvement in sleep before and after taking the placebo for one week varied among the test participants.

4) Selection of subjects Among the test participants, the top 25% (26 people) of test participants who had the highest degree of improvement in OSA sleep questionnaire scores were selected as those with a high placebo effect on sleep improvement (high placebo effect group), and the bottom 25% (26 people) of test participants who had the lowest degree of improvement in OSA sleep questionnaire scores were selected as those with a low placebo effect on sleep improvement (low placebo effect group). A total of 52 test participants from both groups were selected as subjects, and the oil blotting films containing sebum collected in 2) were subjected to the following analysis.

5) RNA preparation and sequencing The oil blotting film was cut to an appropriate size, and RNA was extracted using QIAzol Lysis Reagent (Qiagen) according to the attached protocol. Based on the extracted RNA, reverse transcription was performed at 42°C for 90 minutes using SuperScript VILO cDNA Synthesis kit (Life Technologies Japan, Inc.) to synthesize cDNA. The random primers included in the kit were used as primers for the reverse transcription reaction. From the obtained cDNA, a library containing DNA derived from the 20802 gene was prepared by multiplex PCR. Multiplex PCR was performed using Ion AmpliSeqTranscriptome Human Gene Expression Kit (Life Technologies Japan, Inc.) under the conditions of [99°C, 2 minutes → (99°C, 15 seconds → 62°C, 16 minutes) × 20 cycles → 4°C, Hold]. The obtained PCR product was purified with Ampure XP (Beckman Coulter, Inc.), and then buffer reconstitution, digestion of primer sequences, adapter ligation and purification, and amplification were performed to prepare a library. The prepared library was loaded onto an Ion 540 Chip and sequenced using an Ion S5/XL system (Life Technologies Japan, Inc.). Each read sequence obtained by sequencing was genetically mapped using hg19 AmpliSeq Transcriptome ERCC v1, which is a reference sequence of the human genome, to determine the gene from which each read sequence originated.

6) Data analysis and selection of feature genes The read count of each read obtained by sequencing the SSL-derived RNA of the subject obtained in 5) above was used as the expression level data of each RNA. In order to correct the difference in total read count between samples, each read count was converted to an RPM (Reads per million mapped reads) value. However, reads with a read count of less than 10 were treated as missing values. For 3072 types of genes for which expression level data other than missing values was obtained in 90% or more of the subjects, Student's t test was performed between groups (high placebo effect group and low placebo effect group) based on the expression information, and genes with p<0.05 were extracted. As shown in Table 6, 80 types of genes whose expression varies statistically significantly between groups were obtained. The 47 genes in Table 6A had higher expression levels in the high placebo effect group, and therefore were markers (positive markers) with high expression levels in subjects with a high placebo effect on sleep improvement. The 33 genes in Table 6B had lower expression levels in the high placebo effect group, and therefore were markers (negative markers) with low expression levels in subjects with a high placebo effect on sleep improvement. These genes were selected as feature genes to be used in the following prediction model construction.

7) Model Construction The expression data of the feature genes selected in 6) was converted to a logarithmic value of base 2 (Log ₂ (RPM+1) value) by adding an integer 1 to approximate a normal distribution from the RPM value according to the negative binomial distribution. The Log ₂ (RPM+1) value of the expression data of the obtained 80 feature genes was used as an explanatory variable, and whether the placebo effect on sleep improvement was high or low (high placebo effect group or low placebo effect group) was used as the objective variable to construct a prediction model for discriminating between the high placebo effect group and the low placebo effect group. The random forest algorithm was specified as a method in the "caret" package of the R language, and the optimal value of the number of variables (mtry value) used to construct a decision tree once was tuned. The random forest algorithm was executed using the mtry value determined by tuning, and the estimated error rate (OOB error rate) was calculated. As a result, in a model using the expression data of all 80 genes as variables, the OOB error rate was 26.9%, indicating that the genes shown in Table 6 can be used as markers for evaluating the placebo effect on improving sleep in individuals.

Example 2: Construction of a placebo effect prediction model for sleep improvement using genes with high random forest variable importance 1) Data used 3,072 types of genes for which non-missing expression data was obtained for more than 90% of subjects in 6) of Example 1 were used in the following analysis. In order to approximate the normal distribution from the RPM values that follow a negative binomial distribution, the gene expression data was converted to base 2 logarithmic values ( _Log2 (RPM+1) values) by adding an integer 1.

2) Selection of Feature Genes Using the random forest algorithm, the feature genes used to construct the prediction model were selected. The Log ₂ (RPM+1) values of the expression data of the 3072 genes obtained in 1) were used as explanatory variables, and whether the placebo effect on sleep improvement was high or low (high placebo effect group or low placebo effect group) was used as the objective variable to construct a prediction model that distinguishes between the high placebo effect group and the low placebo effect group. The random forest algorithm was specified as a method in the "caret" package of the R language, and the optimal value of the number of variables (mtry value) used to construct a decision tree once was tuned. Using the mtry value determined by tuning, the random forest algorithm was executed, and the top 100 genes in terms of variable importance based on the Gini coefficient were calculated (Table 7). These 100 genes were selected as feature genes to be used to construct the prediction model.

3) Model Construction A prediction model was constructed to distinguish between the high placebo effect group and the low placebo effect group, using the Log ₂ (RPM+1) values of the expression data of the 100 feature genes selected in 2) as explanatory variables, and whether the placebo effect on sleep improvement is high or low (high placebo effect group or low placebo effect group) as the objective variable. The random forest algorithm was specified as a method in the "caret" package of the R language, and the optimal value of the number of variables (mtry value) used to construct a decision tree once was tuned. The random forest algorithm was executed using the mtry value determined by tuning, and the estimated error rate (OOB error rate) was calculated. As a result, in the model when the expression data of 100 genes was used as a variable, the OOB error rate was 17.3%, and it was shown that the genes shown in Table 7 can be used as markers for evaluating the placebo effect in improving sleep for individuals.

Example 3 Construction of a placebo effect prediction model based on feature genes used in common in Examples 1 and 2 1) Selection of feature genes Among the feature genes used in Examples 1 and 2, the 24 genes shown in Table 8 were used in common in both Examples. These 24 genes were selected as feature genes to be used in constructing a prediction model.

2) Model Construction A prediction model was constructed to distinguish between the high placebo effect group and the low placebo effect group, using the Log ₂ (RPM+1) values of the expression data of the 24 feature genes selected in 1) as explanatory variables, and whether the placebo effect on sleep improvement is high or low (high placebo effect group or low placebo effect group) as the objective variable. The random forest algorithm was specified as a method in the "caret" package of the R language, and the optimal value of the number of variables (mtry value) used to construct a decision tree once was tuned. The random forest algorithm was executed using the mtry value determined by tuning, and the estimated error rate (OOB error rate) was calculated. As a result, in the model when the expression data of the 24 genes was used as a variable, the OOB error rate was 28.9%, and it was shown that the genes shown in Table 8 can be used as markers for evaluating the placebo effect in improving sleep for individuals.

Example 4 Construction of a placebo effect prediction model for sleep improvement based on all feature genes used in any of Examples 1 and 2 1) Selection of feature genes The feature genes used in any of Examples 1 and 2 (a total of 156 types of genes shown in Table 1) were selected as feature genes to be used in constructing a prediction model.

2) Model Construction A prediction model was constructed to distinguish between the high placebo effect group and the low placebo effect group, using the Log ₂ (RPM+1) values of the expression data of the 156 feature genes selected in 1) as explanatory variables, and whether the placebo effect on sleep improvement is high or low (high placebo effect group or low placebo effect group) as the objective variable. The random forest algorithm was specified as a method in the "caret" package of the R language, and the optimal value of the number of variables (mtry value) used to construct a decision tree once was tuned. The random forest algorithm was executed using the mtry value determined by tuning, and the estimated error rate (OOB error rate) was calculated. As a result, in the model when the expression data of 156 genes was used as a variable, the OOB error rate was 28.9%, and it was shown that the above 156 genes can be used as a marker for evaluating the placebo effect on sleep improvement in individuals.

Claims (18)

A method for assessing a placebo effect on improving sleep in a subject, comprising measuring expression of a marker for assessing the placebo effect on improving sleep in the subject;
The marker comprises at least one selected from the group consisting of a nucleic acid encoding a TNK2 gene and a protein encoded by the gene.
method. The method according to claim 1, wherein the marker further comprises at least one selected from the group consisting of nucleic acids encoding the genes shown in Table 1 below and proteins encoded by the genes.

The method according to claim 1, wherein the marker further comprises at least one selected from the group consisting of nucleic acids encoding the genes shown in Table 2 below and proteins encoded by the genes.

The method according to claim 1, wherein the marker further comprises at least one selected from the group consisting of nucleic acids encoding the genes shown in Table 3 below and proteins encoded by the genes.

The method according to any one of claims 1 to 4, wherein the marker is a nucleic acid marker. The method of claim 5, wherein the nucleic acid is mRNA extracted from lipids on the skin surface. The method according to any one of claims 1 to 6, wherein the subject is in a state in which the placebo has not been administered. further comprising assessing a placebo effect on improving the subject's sleep using the predictive model;
the prediction model is constructed by machine learning using the expression level of the marker in each person in a teacher sample population as an explanatory variable and whether the placebo effect in each person in the population is high or low as a response variable;
The teacher sample population includes those with high and low placebo effects on sleep improvement;
The method according to any one of claims 1 to 7. The method according to any one of claims 1 to 7, further comprising evaluating a placebo effect on improving the subject's sleep based on the expression level of the marker. A method for selecting a subject who does not have a high placebo effect on improving sleep, comprising measuring the expression of a marker for evaluating the placebo effect on improving sleep in the subject;
The marker comprises at least one selected from the group consisting of a nucleic acid encoding a TNK2 gene and a protein encoded by the gene.
method. The method according to claim 10, wherein the marker further comprises at least one selected from the group consisting of nucleic acids encoding the genes shown in Table 4 below, and proteins encoded by the genes.

The method according to claim 10, wherein the marker further comprises at least one selected from the group consisting of nucleic acids encoding the genes shown in Table 5 below, and proteins encoded by the genes.

The method according to claim 10, wherein the marker further comprises at least one selected from the group consisting of nucleic acids encoding the genes shown in Table 6 below, and proteins encoded by the genes.

The method according to any one of claims 10 to 13, wherein the marker is a nucleic acid marker. The method of claim 14, wherein the nucleic acid is mRNA extracted from lipids on the skin surface. The method according to any one of claims 10 to 15, wherein the subject is in a state in which the placebo is not administered. Further comprising using the predictive model to select subjects who do not have a high placebo effect on improving sleep;
the prediction model is constructed by machine learning using the expression level of the marker in each person in a teacher sample population as an explanatory variable and whether the placebo effect in each person in the population is high or low as a response variable;
The teacher sample population includes those with high and low placebo effects on sleep improvement;
The method according to any one of claims 10 to 16. The method according to any one of claims 10 to 16, further comprising selecting subjects for whom the placebo effect on improving sleep is not high based on the expression level of the marker.